All modern day graphics cards basically need to be installed into a PCIe (PCI-Express) slot on the motherboard. The word PCIe stands for Peripheral Component Interconnect Express. It’s an upgrade over Accelerated Graphics Port (AGP), which was used by graphics cards before PCIe replaced it.

PCIe has seen a lot of updates and revisions throughout the years. A main feature of it has been that almost every new PCIe version has doubled the bandwidth from its previous version.

Released in 2003, PCIe 1.0 came with a 2.5 gigatransfers per second (GT/s) speed. PCIe 2.0 was released on 2007 and with a speed of 5.0 GT/s. Then came PCIe 3.0 in 2010 which allows speeds up-to 8.0 GT/s. PCIe 4.0, which later released in 2017 again doubled it to 16 GT/s. Then came PCIe 5.0 in 2019 with took it to 32 GT/s.

Newer PCIe versions require the motherboards come with its support. However, all newer versions are backwards compatible. Meaning one can install a new graphics card into a PCIe 1.0 slot and still make it work possibly, though with performance issues.

PCIe uses up-to x16 lanes for the cards to use. It supports x1, x2, x4, x8 and x16 lanes. The more the lanes the more data can be transmitted. The lower one goes, the half the speed. In simple words, a PCIe 5.0 slot running at x8 will perform the same as PCIe 4.0 at x16. Similarly, PCIe 4.0 at 8x will perform same as PCIe 3.0 x16. This same idea continues further as one goes further.

Almost all motherboards support x16 lanes. The problem happens when some graphics cards don’t support them. Especially some really cheaper ones.

While PCIe 6.0 specs are released already, we are yet to see a PCIe 5.0 based gaming graphics card. As Nvidia, AMD and even Intel graphics cards don’t make a use of it and use PCIe 4.0 instead.

The question arises. Are newer PCIe version really useful. Latest gaming benchmarks say otherwise.

PCIe 4.0 and 3.0 Gaming Benchmarks

Simply put. Usually, gaming benchmarks testing the PCIe versions are hard to come by. However, a South Korea based site Quasarzone not only did its testing (translated) but went into detail about the tests.

It tested 5 different games on these PCIe versions:

• PCIe 4.0 / 16x (RTX 4090 standard)
• PCIe 4.0 / 8x
• PCIe 4.0 / 4x
• PCIe 3.0 / 16x
• PCIe 3.0 / 8x
• PCIe 3.0 / 4x

One might find that PCIe 5.0 is missing from this list. That’s because, as mentioned previously, no graphics card comes with it.

The setup it used to test these included the AMD Ryzen 7 5800X3D CPU, DDR4 16 GB 3200 MHz RAM, a WD SN570 M.2 1TB SSD among others. The most important thing, however, is that they did the benchmark on Nvidia RTX 4090 FE, which is a top of the line graphics card and possibly the best option to make the best use of the PCIe 4.0 bus.

Out of those 5 different games which were all tested on max quality, two were additionally tested with Nvidia’s latest DLSS 3.0 Quality setting enabled. Out of those two, one had a limit of 250 FPS, hence raytracing was used in it too at max settings. While there’s a discrepancy in testing, it seems to have been intended to make the max use of the PCIe bandwidth available.

The results were interesting.

Putting PCIe To Gaming Benchmarks

As one can see. At 1080p, the benchmarks showed that there was just a 2-3% decrease in average and 1% low frames per second in PCIe 3.0 x16 when compared to PCIe 4.0 x16. The results were similar in the 4K tests too, where PCIe 3.0 x16 again saw just 2-3% decrease in both average and 1% FPS.

What’s interesting is that individually, almost all these games had almost negligible difference in FPS between PCIe versions. At most 5% in 1% low and 2.5% in average at max. The only exception was one game with DLSS 3.0 and raytracing enabled, where the difference was 10% at 1% low and 5% in average FPS. If we remove one game from the list, the difference between PCIe versions is so small it’s not even worth consideration.

Major difference were however seen when switching to PCIe 3.0 x8 and below. Which basically means running the card at PCIe 2.0 x16, which almost no one uses now.

Frametime

In gaming, one of the often overlooked stat is the frametime. Digital Trends has a good article explaining it. Basically, the lesser the ms in frametime, the better it is. Quasarzone checked frametimes in these gaming benchmarks too. While it did not provide a 5 game average, the results were similar throughout the games except the one mentioned previously (pic different).

In frametimes, by and large, in almost all games it was seen that both PCIe 4.0 and 3.0 performed similarly in different lanes. The exception was PCIe 3.0 at x4 lanes, which is akin to a PCIe 1.0 slot running at x16.

Conclusion

When AMD announced it’s Zen 4 Ryzen 7000 series of processors, it insisted on its PCIe 5.0 capability. When Intel released Alder Lake and Raptor Lake CPUs, they came with a support for PCIe 5.0 too. However, there are no graphics cards available to make a use of PCIe 5.0. This PCIe 5.0 support, though future proof, needlessly make the top motherboards more expensive.

But forget about PCIe 5.0. Graphics cards might not be even utilizing PCIe 4.0 fully. If the above results are to be believed, there could similarly be just 3% difference between PCIe 5.0 and 4.0 in (gaming) benchmarks too. While it’s true that only 5 games were tested and more heavy games need to be tested, but this gives us a fair idea about things.

What’s PCIe 5.0 is useful for, however, is SSDs which can actually make the use of the bandwidth available in it. But in SSDs too there are some doubts, like whether there are any changes visible in newer PCIe SSDs or are they actually required for the common users.

PCIe 4.0 vs 3.0 Shows Only 3% Difference In Gaming Benchmark

Being the world's No. 1 foundry, TSMC was almost immune to the softening high-tech market in Q2 and Q3, but it looks like the party is over even for the world's largest contract chip manufacturer. TSMC's leading customers began to revise their orders to the company, affecting the foundry's results starting from the fourth quarter of 2022, reports DigiTimes. 

According to the report, virtually all TSMC clients will experience a downturn and have to cut orders, so TSMC's utilization will decline significantly in Q1 2023. For example, the utilization rate of TSMC's N7-capable lines (7nm, 6nm-class technologies) will decline to around 50% in early 2023. Furthermore, even TSMC's N5/N4-capable lines will be underutilized, though this may not come as a surprise since these are used to make leading-edge products, like Apple's smartphone SoCs, and demand for advanced handsets is usually down in the first half of the year. More alarming is that even N28-capable fabs — which have been fully loaded since the beginning of the chip deficit in early 2021 — will be underutilized. 

Due to the slowing economy in China as well as its COVID lockdowns, an economic downturn in numerous European countries, and reduced demand for many products in the U.S., large computer hardware, PCs, and smartphone makers lowered their procurement of new chips from companies like AMD, Intel, MediaTek, and Nvidia. As a result, the fabless chip designers reportedly had no choice but to slash orders to TSMC. 

The order cuts will come into effect in Q4 2022, which will increase TSMC's inventories, though it is unclear how the significantly reduced orders will affect the company's revenue this quarter. Meanwhile, DigiTimes estimates that TSMC's Q1 2023 sales will decrease by 15% quarter-over-quarter. By contrast, TSMC's Q1 2022 revenue exceeded its Q4 2021 revenue by 12.1%. 

Slashing orders to a contract manufacturer is not trivial since fabless chip designers are obliged to procure a fixed number of wafers in certain quarters. Nevertheless, TSMC is reportedly willing to accept compensation (as it will hold wafers with chips from AMD, Intel, Nvidia, etc., before they are ready to buy them) and even renegotiate deals on long-term supply contracts (i.e., increase the number of wafers that a company is committed to buying in the future) in exchange. Such measures will not make TSMC's life any easier in Q4 2022 or Q2 2023, though. 

Market observers generally remain optimistic about the demand for advanced chips returning to normal in 2023. As a result, TSMC is still expected to post revenue growth for the year, but the sales increase may not be as impressive as the 2021 – 2022 period.

Source

SEOUL – South Korean President Yoon Suk-yeol urged additional tax breaks for the domestic semiconductor industry as the nation’s flagship export sector reels from a perfect storm – one that continues to rise in intensity.

The semiconductor super-cycle, which buoyed the Korean economy through the Covid-19 pandemic, turned south this summer as chip shortages turned to gluts and prices plunged. Even more ominously, the future geography of the sector looks set for a seismic shift as the US applies increasing pressure on its allies’ chip businesses with China.

In addition to global factors such as the war in Ukraine, related commodity inflation and rising interest rates, this combination of economic pressures and unprecedented uncertainty hanging over the flagship national export is mirrored in the hideous performance of Korean stocks in 2022.

The benchmark KOSPI index opened to a bullish 2,999.75 points on December 30, 2021.. This Friday, 2022’s last trading day, it closed at 2,236.4 points.

Yet despite its massive implications, there is still no clear answer to the huge question posed by US policy toward China and semiconductors as Seoul grapples with its impossible positioning: It balances between top trade partner Beijing and leading strategic ally Washington.

But even as chip executives fret over future strategies for investment in, and sales to, China, if politicians decide to press ahead with decoupling from Chinese tech, they may be pushing on an open door.

A September global public opinion poll of 56 nations delivered the surprise news that South Koreans harbor stronger anti-China sentiment than any people on Earth.

Washington is clearly piling heavy-handed trade pressure upon Seoul, and huge anti-American riots shook South Korea in 2002 after the death of two schoolgirls in a road accident with GIs. Yet in the September survey, the United States emerged as the most popular nation among South Koreans.

South Koreans hold an American flag at a conservative political rally in a file photo. Image: Twitter

All for chips

Yoon, speaking to officials on Friday morning, reinforced the centrality of the semiconductor industry to the Korean economy.

“Strategic technologies, such as semiconductors, are a national-security asset and our industries’ core technology, so I would like the finance ministry to actively consider ways to additionally expand tax breaks for national strategic industries, including the semiconductor industry, in consultation with relevant ministries,” he said.

Corporations can currently bag an 8% break on semiconductor investments thanks to a recent revision to the previous tax code, which previously offered a 6% break. However, that 2-percentage-point rise falls far short of the whopping 20% tax break that Yoon’s conservative People Power Party had called for.

The National Assembly is controlled by the progressive Democratic Party.

South Korea is home to the world’s two leading memory chip manufacturers – respectively, Samsung Electronics and SK hynix. Facing the cyclical downturn, they are taking different approaches to their capital expenditures in 2023.

Sector leader Samsung, damning the torpedoes, is ramping up production at its new fab in the Yellow Sea port of Pyeongtaek – the world’s largest chipmaking facility – by 10%. SK hynix, on the other hand, is following other industry players in cutting back its capex for the year – by a whopping 50%.

Yet while the peaks and troughs of the chip cycle are known challenges that related businesses are experienced in managing, a darker shadow looms: the rising demands from the US that its allies, from Western Europe to East Asia, cease commercial cooperation with China on cutting-edge chips.

Backing up this demand, the weapon the US could deploy is its multiplicity of patents in high-end chip design suites and manufacturing equipment.

Exiting Chinese investments would be agonizing for South Korea. According to local media, Samsung makes some 40% of its total NAND memory in Xian, while SK hynix makes 50% of its DRAM memory in Wuxi.

So too would cutting trade. In 2021, according to data collated by bank Santander, China was South Korea’s leading trade partner: 27% of Korea’s total exports went to the mainland, with another 6.2% going to Hong Kong. The United States was in distant second place, taking some 15% of Korea’s overseas shipments.

It’s the same story in terms of supplying imports: China accounts for 24%, with Hong Kong adding 6.2%; the US, again in second place, supplies just 13%.

In Korea’s overall export portfolio, chips and chip parts accounted for 16.2% of the total, making them the leading item, Santander found. The next-largest export item, autos, took up just 7.0%.

Leading academic Moon Chung-in, who has advised three different South Korean administrations on international relations, laid out the massive reliance Korea’s chip sector has on China.

Historically, Korean chipmakers “export about 60% of their products to China (40% to the Chinese mainland, 20% to Hong Kong) while importing almost 60% of chip-related critical materials from China,” Moon wrote on these pages. 

In 2021, the last year for which total data are available, mainland China was the destination for 39.7% of Korea’s semiconductor exports, a higher ratio than China’s share of Korea’s total exports, which is 25.3%.

There is some breathing space. In October, South Korean chipmakers won a one-year waiver from US authorities. But given the size of their commerce with China, a huge cloud of opacity hangs over what will happen to their China businesses, post-waiver. Absent a surprise turnaround in US policy, Korean companies look trapped.

“Requests from the US, the world’s most powerful country and Korea’s closest ally, must not be disregarded,” the left-leaning Hankyoreh Daily newspaper editorialized in July. “It is clear that anxiety is high in industrial circles.”

In October, the right-leaning Dong-A daily headlined a piece “US aims to thwart Chinese semiconductor industry on all fronts.”

Given this angst, why is more public noise noise not being aired on the conundrum? Perhaps because silence is golden.

“The Americans are saying, ‘Why don’t you sacrifice your business for our politics?’ and the Koreans don’t want to do it,” said Scott Foster, a Tokyo-based tech analyst with Lightstream Research. “But they are caught between a rock and a hard place due to the US defense alliance, so they want to avoid the question.”

US President Joe Biden and his South Korean counterpart Yoon Suk-youl leave after a joint news conference at the People’s House in Seoul, South Korea, on May 21, 2022. Photo: Pool / Screengrab / NTV

Leaning West

Signs are emerging that Korean companies have already made their choice.

Samsung has been cutting back on its China investments for years. SK hynix remains heavily invested but admitted to investors in a conference call in October that it is committed to diversifying manufacturing and could, if necessary, divest its Chinese fabs.

Meanwhile, both are hedging their bets, spending massively on new plants in the US, where Washington is not only applying political and trade pressure, it is also dangling the carrot of investment incentives.

In a July videoconference with US President Joe Biden, SK Group chairman Chey Tae-won announced $22 billion in investment plans in the USA, including $15 billion in chipmaking.

In 2021, Samsung announced $17 billion in additional investments in its Texas semiconductor manufacturing operations. In July this year, documents filed with Texas authorities showed the company was mulling building 11 more plans, with a value of $200 billion, in the US state over the next two decades.

And the Koreans are not alone. Their regional competitors are hurtling stateside too.

“The amount of investment going into the States from Japan, South Korea and Taiwan is absolutely enormous,” Foster said. “The China thing has given these companies a huge opportunity to grab a big chunk of the US market, from the inside, without the protectionism. Americans are not going to build those factories.”

This could signal one answer to the problem that South Korean – and other Asia chipmakers – face. With free trade now a dead duck, and multiple nations and blocs seeking sovereign chip-production facilities, there is a possibility of mirror sectors rising in different geographies.

“It is getting that everybody wants to build their own industry,” said Foster – a factor that presents its own downside. “There will not be a shortage of semiconductors – there will be permanent overcapacity!” he added.

The question of how much influence soft – and arguably fickle – factors like public opinion wield over hard factors like real-world economics may be debated. Nevertheless, signals indicate that South Korean CEOs have chosen a side – the US – that is in line with public sentiment.

The economic importance of densely populated, prosperous and proximate China is a fact well known to the South Korean public. It is likewise known that, increasingly, the United States is leveraging its strategic alliance with South Korea to pressure China.

Yet a range of public opinion polls in the last two years have shown that South Korean sentiments toward China are plummeting.

Transnational air pollution is one of several reasons Koreans are turning anti-China. A worker walks by the main gate of a coal-to-oil plant in Changzhi in Shanxi province. Photo: Fred Dufour / AFP

‘No thanks’ China; ‘yes please’ America

The Sinophone Borderlands project is a series of online public opinion surveys taken between 2020 and 2022, investigating global attitudes toward China and vice versa. The body comprises academics across the world, including in China itself.

The project’s September online survey of 1,363 Koreans, conducted by the Central European Institute of Asia Studies, found that 81% of South Koreans have negative or very negative views of China – more than any other of the 56 states surveyed.

The figure puts Korea well ahead of the No 2 ranked nation, Switzerland, where 72% of persons hold negative/very negative views, and No 3 ranked Japan, with 69%.

In a detailed analysis, two of the report’s authors noted that South Koreans, if asked to choose between China and the US, held very, very clear views. Some 4% of Koreans were undecided, 5% chose “China” – and a whopping 91% said they would side with the US.

Japan, customarily the most despised country among South Koreans, dropped to fourth place, behind China, Russia and North Korea.

And while there are certainly leftist and even anti-American sentiments simmering within the Korean body politic, the survey found less than 20% of Koreans held negative or very negative views of the US.

There is more. In economic terms, more than 40% of Koreans considered the US “very strong,” while the same figure for China was less than 10%.

What is behind this massive dislike of, and disrespect for, China?

The two countries had a serious spat in 2017. When Seoul agreed to the deployment of a US-manned THAAD (terminal high altitude area defense) anti-missile battery on South Korean soil as a defense against North Korean projectiles, Beijing, insisting that the system’s radars could spy on China, went ballistic.

In retaliation, protests impacted the businesses of South Korean conglomerates in China, leading to dwindling sales by Hyundai Motor, and Lotte withdrawing its retail operations from the market. Chinese tour groups ceased visiting South Korea, dealing a heavy blow to the tourism-focused island of Jeju. Even imports of “K-culture” – pop music, TV dramas, films and computer games – were halted.

Nothing so dramatic has happened since, but the reputation of China, in Korean eyes, has continued to deteriorate.

The study authors found Koreans were angry with China over matters as diverse as transnational air pollution and environmental destruction; Covid-19’s origin; dictatorial communism; and cultural appropriation. The last one is highly visible in wars of words between netizens over aspects of traditional heritage.

Citing past business trips to South Korea, Foster expressed surprise at the apparent shift in public opinions.

“In the past, it was older Koreans who used to be worried about China and North Korea, and the younger ones, worried about their jobs, and it was, ‘US get lost,’” he said. “If that has changed, that is big deal.”

Source

According to DigiTimes, TSMC's utilization rate is expected to experience a significant drop in the first quarter of the year. This is reflected in the stacking of wafer banks, which has reached a new high, and a 15% reduction in orders from leading customers. Almost all of TSMC's clients are expected to experience a downturn and will need to reduce orders in the first quarter of 2023, leading to a significant decline in TSMC's utilization rate. This will affect all of TSMC's production lines, including those that use 7nm, 6nm-class technologies (N7-capable lines), which are expected to have a utilization rate of around 50% in early 2023.

TSMC's N5/N4-capable lines, which are typically used to produce advanced products like smartphone SoCs, are expected to be underutilized. The report also indicates that even TSMC's N28-capable fabs, which have been fully loaded since the beginning of the chip deficit in early 2021, will experience underutilization. The decrease in demand for advanced handsets in the first half of the year is likely contributing to this trend as popular products like the iPhone usually launch towards the end of the year.

Several factors, including a slowing economy in China due to COVID-related lockdowns and reduced demand for many products worldwide, have led to a decrease in the procurement of new chips from companies like AMD, Intel, MediaTek, and Nvidia by large computer hardware, PC, and smartphone manufacturers. As a result, fabless chip designers have reportedly been forced to cut their orders to TSMC.

The reported cuts in orders for TSMC are expected to take effect in the fourth quarter of 2022, which will lead to an increase in the company's inventories. It is not yet clear how these significantly reduced orders will impact TSMC's revenue for this quarter. However, DigiTimes estimates that TSMC's sales for the first quarter of 2023 will decrease by 15% quarter-over-quarter. This is in contrast to the first quarter of 2022, when TSMC's revenue exceeded its revenue for the fourth quarter of 2021 by 12.1%.

Source

As it was released earlier, it makes sense to start with GPT-3 (Generative Pre-trained Transformer 3). Developed by OpenAI, GPT-3 is a general-purpose language model that can generate human-like text based on user prompts and perform a wide range of related language tasks, such as translation, summarization of long texts into more easily digestible points, question answering, and more. The large language model is trained on a diverse dataset of billions of words and has the ability to generate coherent human-like paragraphs of text.

ChatGPT is a variant of the GPT-3 language model also developed by OpenAI. Both are similar in that ChatGPT and GPT-3 are based on the Transformer architecture, which makes it into the name of the two tools and both use large amounts of data and powerful computational resources to learn the patterns and structure of language, which they can then generate coherently. However, there are differences between the two that are worth noting.

The major difference is that the GPT-3 protocol is much larger than ChatGPT. The former has a whopping 175 billion parameters making it one of the largest and most powerful AI language processing models to date, while the latter has 20 billion parameters. This is because ChatGPT has been specifically designed to generate human-like text in the context of chatbot conversations. Accordingly, it has been trained with specific datasets of chatbot interactions. This is likely why one of the reasons there are so fewer parameters.

The other big difference between the two is accessibility. ChatGPT is much more readily available to the public while GPT-3 is reserved for more considered use. This is why ChatGPT has exploded in popularity compared to a more limited awareness among the general public GPT-3.

In short, then, ChatGPT is a more specialized version of GPT-3, which has been optimized for more chat-like interactions. This makes it much more fun to use. On top of this it is easier to access and free to use compared to its older sibling. If you would like to check out ChatGPT and give it a go, check out our ChatGPT guide, which will explain in greater detail what is, how to use it, and what to use it for.

Source

The U.S. Department of Energy said on Dec. 13, 2022, that for the first time – and after several decades of trying – scientists have managed to get more energy out of the process than they had to put in.

But just how significant is the development? And how far off is the long-sought dream of fusion providing abundant, clean energy? Carolyn Kuranz, an associate professor of nuclear engineering at the University of Michigan who has worked at the facility that just broke the fusion record, helps explain this new result.

c

Fusion is the same process that powers the Sun. NASA/Wikimedia Commons

What happened in the fusion chamber?

Fusion is a nuclear reaction that combines two atoms to create one or more new atoms with slightly less total mass. The difference in mass is released as energy, as described by Einstein’s famous equation, E = mc2 , where energy equals mass times the speed of light squared. Since the speed of light is enormous, converting just a tiny amount of mass into energy – like what happens in fusion – produces a similarly enormous amount of energy.

Researchers at the U.S. Government’s National Ignition Facility in California have demonstrated, for the first time, what is known as “fusion ignition.” Ignition is when a fusion reaction produces more energy than is being put into the reaction from an outside source and becomes self-sustaining.

The fuel is held in a tiny canister designed to keep the reaction as free from contaminants as possible. U.S. Department of Energy/Lawrence Livermore National Laboratory

The technique used at the National Ignition Facility involved shooting 192 lasers at a 0.04 inch (1 mm) pellet of fuel made of deuterium and tritium – two versions of the element hydrogen with extra neutrons – placed in a gold canister. When the lasers hit the canister, they produce X-rays that heat and compress the fuel pellet to about 20 times the density of lead and to more than 5 million degrees Fahrenheit (3 million Celsius) – about 100 times hotter than the surface of the Sun. If you can maintain these conditions for a long enough time, the fuel will fuse and release energy.

The fuel and canister get vaporized within a few billionths of a second during the experiment. Researchers then hope their equipment survived the heat and accurately measured the energy released by the fusion reaction.

So what did they accomplish?

To assess the success of a fusion experiment, physicists look at the ratio between the energy released from the process of fusion and the amount of energy within the lasers. This ratio is called gain.

Anything above a gain of 1 means that the fusion process released more energy than the lasers delivered.

On Dec. 5, 2022, the National Ignition Facility shot a pellet of fuel with 2 million joules of laser energy – about the amount of power it takes to run a hair dryer for 15 minutes – all contained within a few billionths of a second. This triggered a fusion reaction that released 3 million joules. That is a gain of about 1.5, smashing the previous record of a gain of 0.7 achieved by the facility in August 2021.

How big a deal is this result?

Fusion energy has been the “holy grail” of energy production for nearly half a century. While a gain of 1.5 is, I believe, a truly historic scientific breakthrough, there is still a long way to go before fusion is a viable energy source.

While the laser energy of 2 million joules was less than the fusion yield of 3 million joules, it took the facility nearly 300 million joules to produce the lasers used in this experiment. This result has shown that fusion ignition is possible, but it will take a lot of work to improve the efficiency to the point where fusion can provide a net positive energy return when taking into consideration the entire end-to-end system, not just a single interaction between the lasers and the fuel.

Machinery used to create the powerful lasers, like these pre-amplifiers, currently requires a lot more energy than the lasers themselves produce. Lawrence Livermore National Laboratory, CC BY-SA

What needs to be improved?

There are a number of pieces of the fusion puzzle that scientists have been steadily improving for decades to produce this result, and further work can make this process more efficient.

First, lasers were only invented in 1960. When the U.S. government completed construction of the National Ignition Facility in 2009, it was the most powerful laser facility in the world, able to deliver 1 million joules of energy to a target. The 2 million joules it produces today is 50 times more energetic than the next most powerful laser on Earth. More powerful lasers and less energy-intensive ways to produce those powerful lasers could greatly improve the overall efficiency of the system.

Fusion conditions are very challenging to sustain, and any small imperfection in the capsule or fuel can increase the energy requirement and decrease efficiency. Scientists have made a lot of progress to more efficiently transfer energy from the laser to the canister and the X-ray radiation from the canister to the fuel capsule, but currently only about 10% to 30% of the total laser energy is transferred to the canister and to the fuel.

Finally, while one part of the fuel, deuterium, is naturally abundant in sea water, tritium is much rarer. Fusion itself actually produces tritium, so researchers are hoping to develop ways of harvesting this tritium directly. In the meantime, there are other methods available to produce the needed fuel.

These and other scientific, technological and engineering hurdles will need to be overcome before fusion will produce electricity for your home. Work will also need to be done to bring the cost of a fusion power plant well down from the US$3.5 billion of the National Ignition Facility. These steps will require significant investment from both the federal government and private industry.

It’s worth noting that there is a global race around fusion, with many other labs around the world pursuing different techniques. But with the new result from the National Ignition Facility, the world has, for the first time, seen evidence that the dream of fusion is achievable.

Source

The move comes as Twitter is facing eviction from Seattle’s Century Square Tower since the company has stopped paying rent, sources familiar with the matter told The New York Times.

Tech news publication Platformer reported that those who worked at the Seattle office have been told to work remotely, according to an email sent to employees.

Twitter will likely now only have offices in New York City and San Francisco, according to the report. However, the company has also reportedly been skipping rent at its Bay Area headquarters.

According to the Times report, the company had cut janitorial and security services at the Seattle office, leaving some employees to bring their own toilet paper to work.

In November, 208 workers at the Seattle offices lost their jobs after Musk slashed Twitter’s 7,500-person workforce in half, according to The Seattle Times.

Workers at the Seattle branch had been affected by Musk’s layoffs in November.
Getty Images

Since Musk acquired Twitter for $44 billion, its workforce is down 75%, according to The New York Times.

Musk said earlier this month he plans to step down as CEO once he finds someone to replace him.

Source

Musk’s scramble to save money included a move this month to abruptly fire janitors who worked at the company’s San Francisco headquarters. The cuts came after the janitors had gone on strike for more pay.

The absence of janitors has “left the office in disarray,” with bathrooms that have “grown dirty” and persistent smells of “leftover takeout food and body odor,” the New York Times reported, citing sources familiar with the matter.

The report said some employees are bringing in their own toilet paper because no janitors are on hand to replace supplies. Musk has purportedly packed Twitter employees onto two floors while shutting down four others.

Other budget-conscious moves included a decision by Musk to shut down one of Twitter’s data centers in Sacramento, Calif., despite reported concerns among some employees that it could hurt the site’s performance.

On Christmas Eve, Musk tweeted that Twitter was still working “even after I disconnected one of the more sensitive server racks.” The site experienced a widespread outage on Thursday, though its exact cause remains unclear.

Elon Musk is Twitter’s owner and CEO.
ullstein bild via Getty Images

The report said Twitter has missed rent payments on its San Francisco offices and stopped paying rent entirely at its building in Seattle, where it now faces evictions. The company also axed cleaners and some security staff at one of its buildings in New York.

Additionally, the Musk-led company has reportedly missed or delayed payments to some vendors, including account firm KPMG and benefits provider Carrot. Twitter’s new leaders are said to be aiming to renegotiate some of the agreements to lower costs.

According to the New York Times, Musk has turned off some Twitter employees with an “erratic and hands-on style” of leadership. The Twitter CEO has directed his subordinates to identify employees responsible for leaking details about the company’s operations to the press.

In all, Musk is reportedly attempting to shave $500 million in nonlabor expenses from Twitter’s budget, according to the outlet.

Elon Musk has reportedly consolidated San Francisco workers on two floors.
Anadolu Agency via Getty Images

The billionaire has been open about Twitter’s financial issues since he bought the company for $44 billion in late October. Earlier this month, Musk argued his various cuts were necessary to avoid a $3 billion budgetary shortfall.

“That is why I spent the last five weeks cutting costs like crazy,” Musk said during a Twitter Spaces event last week. “This company is like, basically, you are in a plane that is headed toward the ground at high speed with the engines on fire and the controls don’t work.”

The Post has reached out to Twitter for comment. The company has cut its entire communications team.

Source

An overview of how the ChatGPT NLP model works

What is ChatGPT exactly?

We already know that ChatGPT is a chatbot, but it is not like an ordinary chatbot you interact with when, for example, dealing with an online customer service bot. So, how exactly does it work?

ChatGPT is based on the GPT-3 language model, a technology created by OpenAI. GPT stands for Generative Pre-trained Transformer model and the 3 is there because it is the third generation of the tool. As the name suggests, the tool generates text as a response to user’s text prompt, using the massive amounts of training data that OpenAI have used to build the model. Therefore, like other types of generative AI such as Stable Diffusion or M idjourney, ChatGPT has been trained with huge amounts of data, which in this specific case are written texts.

However, with text being text, ChatGPT’s possibilities go beyond the written word and can go into more complex areas of text use including coding and programming, translations, reports, and even more “human” aspects of conversational interactions such as offering therapeutic responses to any of life’s problems you may think to throw at it.

In fact, ChatGPT’s answers will likely surprise you because of how elaborate they can end up. However, the tool is still far from perfect and as we’ve covered in previous articles (see links above) the tool can’t be relied upon for factual accuracy. ChatGPT does not have an Internet connection and it can often give inaccurate responses or erroneous statements. Therefore, it is important that you check all the information it gives you before you use it for any serious purpose.

Even so, despite these types of errors and failures, ChatGPT has proven itself to be a truly impressive tool with many internet users openly speculating about all the weird and wonderful ways it will revolutionize the world.

How do I use ChatGPT?

A key aspect of its success stems from the fact that using ChatGPT is not a complicated affair. To start, you will have to access its official website, chat.openai.com. Once inside, OpenAI will prompt you to sign up and create an account before you start using the.

For now, at least, ChatGPT is completely free, and you will not have to incur any additional costs to use it. When you are done with the registration process, OpenAI will take you to the main page of ChatGPT chatbot, which is a window that looks just like an ordinary messaging app. Here you will see a text box where you will have to write your questions or prompts.

You can then begin asking ChatGPT things and it is really as simple as that. Of course you have to remember that isn’t connected to the internet so it won’t be able to answer everything and that it is a natural language processing model and not a generally intelligent machine. These two limitations set the technical boundaries for what you can ask it and what it is capable of asking. Other limitations relate to how these types of tools are created as well as how they are intended to be used. These tools are built using vast amounts of data, which means any problems in the data, for example, if it includes biased opinions, could lead to problematic responses from the tool. OpenAI tries to stop this from happening through the use of filters, which you may come up against if they work as they are supposed to. Also, ChatGPT is not supposed to create any NSFW content so any attempts to have it do so will come up against blocks too.

Of course, the other thing you should keep in mind is that everything you write will be registered on the OpenAI database. Therefore, try not to type in any personal data or valuable information.

What can I do with ChatGPT?

Today, ChatGPT is a technology does not have a specific use, which is probably why so many people have been going so crazy about it. It can do a lot of different things, which means you have wide variety of potential use cases to explore. For example, you could ask it to help you write a script or a play, summarize a book, get into coding and programming, or even write poems.

To get an idea of the many different types of things people have been using for and that you could use it for, check out our first look at ChatGPT, which we put together at the beginning of December.

Source

The "water battery" itself. Image credit: Beketoff/Shutterstock.com

A giant "water battery" in the Swiss Alps is finally operational, 14 years after construction began. The Nant de Drance Hydropower Plant can store enough power to charge 400,000 car batteries.

Until fusion (fingers crossed for the next few decades/centuries) is capable of solving all of our power needs, we have a problem to deal with in renewable energy sources: what to do when more power is produced than we actually need. On particularly windy or sunny days, production from wind and solar farms can outstrip demand for that power. While this is obviously a better problem than the reverse, and sometime even leads to a situation where consumers are paid to consume electricity, storage solutions are needed for days when the sun isn't shining and the wind is still.

The Nant de Drance Hydropower Plant is one novel solution for storing large amounts of this excess energy. It's not exactly a new idea, but it is effective nonetheless. During times when renewable energy sources are producing more energy than the demand, the excess energy is used to pump water from the plant into the Vieux Emosson reservoir above. The water can then be used by the hydropower plant below when more power is needed, using the water to power turbines as it passes through at 360 cubic meters per second.

The idea has been dubbed a "water battery", in that the energy is "stored" as water, just at a slightly higher altitude than it was previously. While the efficiency isn't perfect, it is far, far better than letting the excess power go to waste.

"There are losses like any storage, but the yield is very good," director Alain Sauthier told Reuters. "We have about 80% efficiency over the complete cycle."

“In less than ten minutes we can reverse the direction of rotation of the turbines and switch from electricity production to storage," he added to Swissinfo. "Such flexibility is key in order to react promptly to the needs of the electricity grid and adapt electricity generation and consumption. Otherwise, you risk a collapse of the grid and blackout".

The plant is capable of providing storage for much of Europe, not just Switzerland. Now fully operational, it is one of the most powerful plants on the continent. 

Source

Your smartphone: now in 3D. Image credit: KristianDESIGN/Pixabay 

In 1943, Thomas Watson, the president of IBM, famously predicted the world market for computers would top out at “maybe five” of the machines. He was wrong – you likely have more than that in your own house, let’s face it – but at the time, it made sense. After all, if computers were still gigantic, vacuum-tube-powered addition machines, you probably wouldn’t want more than about five either.

It’s a similar story with holograms. Even back in the 1990s, more than 40 years after Dennis Gabor first came up with the idea of using wavefront interference to reconstruct images in three dimensions, science fiction was still assuming the need for entire decks and suites to power our holographic adventures.

In fact, they can run on a smartphone.

Almost two years ago, researchers at MIT made a breakthrough – a technology they dubbed “tensor holography”. Since then, the project has continued to improve, and today, the team are working with a system they say is “fully automatic, robust to rendered and misaligned real-world inputs, produces realistic depth boundaries, and corrects vision aberrations.”

“We are amazed at how well it performs,” project co-author Wojciech Matusik commented back in 2021. And it’s economical, too: the amount of computer power and memory required for this real-time 3D holography is less than one megabyte. 

“It’s [a] negligible [amount],” Matusik pointed out, “considering the tens and hundreds of gigabytes available on the latest cell phone.”

Holograms have come a long way since the first laser-generated static images of the early 20th century. Even then, though, it was a complicated thing to set up: it required splitting a laser beam in two, with half the beam being used to illuminate the subject and the other half used as a reference for the light waves’ phase. 

Computers made this process easier – but they came with their own problems. Supercomputers designed to run physics-based simulations of the laser setup were imperfect and computationally expensive: “Because each point in the scene has a different depth, you can’t apply the same operations for all of them,” explained study lead Liang Shi. “That increases the complexity significantly.”

So the team took a different tack entirely. This is the future, they reasoned, and we don’t always need to come up with our own solutions for problems – not when we can just set a computer to teach itself. They built a convolutional neural network and set it to match up 4,000 pairs of computer-generated images: one a 2D picture containing information on the color and depth of every individual pixel, the other a hologram of that image.

The result: a computer program so good at creating holograms that it surprised even the team themselves. “It’s a considerable leap that could completely change people’s attitudes toward holography,” Matusik said. “We feel like neural networks were born for this task.”

So where now for the developing technology? Some experts have pointed out the advantages of holograms over the occasionally nauseating and eye-straining virtual reality (VR) – perhaps the holodeck is a more likely future than the metaverse after all. Other applications include 3D printing, the team point out, as well as medical visualizations, microscopy, and materials science.

“Holographic 3D displays provide differentiating interactive experiences from cell phones or stereoscopic augmented reality (AR) and [VR] displays,” the team’s latest paper on the topic explains. “[Our work] runs in real-time on a consumer-grade GPU [graphics processing unit] and 5 FPS [frames per second] on an iPhone 13 Pro...promising real-time mobile performance in future-generation AR/VR headsets and glasses.”

The latest study is published in Light: Science & Applications.

Source

Artist's rendering of OSIRIS-REx spacecraft descending towards Bennu to collect a sample of the asteroid’s surface. Image Credit: NASA/Goddard/University of Arizona

It’s been an eventful year for space exploration, with successes including the completion of Nasa’s Artemis 1 mission (finally), the inauguration of the James Webb Space Telescope, and the completion of China’s Tiangong space station.

2023 is set to be another busy year. Here are five of the most exciting missions to watch out for.

1. Jupiter Icy Moons Explorer

In April, the European Space Agency (Esa) is set to launch the Jupiter Icy Moons Explorer (Juice), in what will be Europe’s first dedicated robotic mission to Jupiter. Juice is due to reach the planet in July 2031 after performing an incredible flight path through the Solar System. The mission will enter into orbit around Jupiter and perform numerous flybys of its large icy moons: Europa, Ganymede and Callisto.

After four years of moon flybys, Juice will then enter into orbit around Ganymede, the largest moon in the Solar System – becoming the first spacecraft ever to reach orbit around the moon of another planet. The icy moons of Jupiter are interesting as they are all believed to host oceans of liquid water beneath their frozen surfaces. Europa, in particular, is regarded as one of the most likely abodes in the Solar System for extra-terrestrial life.

Juice will be equipped with ten scientific instruments including ice-penetrating radar to study the internal oceans. This use of radar is a practical first step in mapping the sub-surface oceans, paving the way for more exotic future missions involving submersible vehicles – some of which have already been put forward. The launch window runs from April 5 to April 25.

2. SpaceX Starship

Although no date has been announced by aerospace company SpaceX at the time of writing, the first orbital test flight of the super-heavy Starship spacecraft is highly anticipated to occur in early 2023. Starship will be the largest spacecraft capable of carrying humans from Earth to destinations in space (the International Space Station is larger, but it was assembled in space). It will be the most powerful launch vehicle ever to fly, capable of lifting 100 tonnes of cargo to low Earth orbit.

Starship is the collective name for a two-component system consisting of the Starship spacecraft (which carries the crew and cargo) and the Super Heavy rocket. The rocket component will lift Starship to some 65km altitude before separating and returning to Earth in a controlled landing. The upper Starship component will then use its own engines to push itself the rest of the way to orbit.

Several short test flights of the Starship portion of the system have been made with varying degrees of success. But the upcoming flight will be the first time the whole system will be used to reach space as one. This first orbital flight was originally scheduled to launch in September 2022, but has been delayed several times.

3. dearMoon

The long-awaited dearMoon project, which will take members of the public on a six-day trip around the Moon and back, is due for launch on Starship and was originally planned for 2023. The exact date will depend on the successful test of Starship, but has been on the books since 2018. It will be the first true deep space tourism launch.

Financed by business entrepreneur Yusaku Maezawa, a competition was set up to select eight members of the public (and an unknown number of crew) to join Maezawa on the trip – all completely paid for. The winners and criteria used have not been disclosed, although it is suspected the guests may be established or aspiring artists.

This mission will mark a big change in the way we think about space, as previously only astronauts picked using incredibly stringent criteria have been able to go into deep space (note: we are not counting brief 10-minute jaunts up to 100 km). A full trip of several days poses extreme risks, both in terms of health and engineering.

The success or failure of the dearMoon mission could affect whether deep space tourism becomes the next big thing, or it is relegated back to being a pipe-dream.

4. Asteroid explorer returns to Earth

The Origins Spectral Interpretation Resource Identification Security – Regolith Explorer, mercifully more commonly known as OSIRIS-REx, is a Nasa mission to near-Earth asteroid Bennu. A key goal of this robotic mission was to acquire samples of Bennu and return them to Earth for analysis.

OSIRIS-REx is now fast returning to Earth with up to a kilogram of precious asteroid samples stored aboard. If all goes well, the capsule will detach from the spacecraft, enter the Earth’s atmosphere and parachute to a soft landing in the deserts of Utah on September 24. Asteroid sample return has only been achieved once before, by the Japanese Space Agency’s Hayabusa 2 mission in 2020.

Bennu the golden space rock. NASA/Goddard/University of Arizona

Bennu is an approximately diamond-shaped world just half a kilometre in size, but has many interesting characteristics. It is believed to have broken off from a much larger asteroid in the first 10 million years of the Solar System. Some of the minerals detected within it have been altered by water, implying that Bennu’s ancient parent body possessed liquid water.

It also has an abundance of precious metals, including gold and platinum. Finally, Bennu is classed as a potentially hazardous object with a (very) small possibility of Earth impact in the next century.

5. India’s private space launch

While SpaceX is the most prominent private space launch company, there are many others developing their own series of launchers around the world. Skyroot Aerospace, which successfully launched its Vikram-S rocket in November 2022, is soon to become the first private Indian company to launch a satellite.

The rocket itself reached 90km in altitude, a distance that would need to be improved upon to get a constellation of satellites into orbit. Skyroot’s first satellite launch is planned for 2023, with a goal of undercutting the cost of private space launch rivals by producing its 3D-printed rockets in a matter of days. If successful, this could also provide a route for cheaper launches of scientific missions, enabling a faster rate of research.

Clearly, interest in the space sector remains high. With many bold advances and launches due in 2023, we are entering a new phase akin to the “Golden era” of space launches in the 1960s and ’70s.

Source

Eyes? Where we're going (the physics lab) we don't need eyes. Image credit: Ezume Images/Shutterstock.com

As a general rule, if you want sight, you need light. You’re only reading this right now thanks to the light from your screen being beamed onto your retinas, converted into electrical signals, and sent up the optic nerve for your brain to interpret as a bunch of words and images.

But what if you could see things without all that rigamarole? It might sound impossible – perhaps even counter to the very definition of sight – but thanks to the bizarre world of quantum mechanics, it’s actually perfectly possible.

“Since the inception of quantum mechanics, the quest to understand measurements has been a rich source of intellectual fascination,” notes a new paper published this month.

“The interaction-free measurements belong to the class of quantum hypothesis testing, where the existence of an event (for example the presence of a target in a region of space) is assessed,” it explains. “Here… the task is to detect the presence of a microwave pulse…  [such] that at the end of the protocol the detector has not irreversibly absorbed the pulse.”

In other words: find a way to “see” a microwave pulse, without using a single photon.

If successful, the Aalto University team behind the new paper wouldn’t be the first to achieve such a feat – in fact, their experiment was based on one originally performed by Anton Zeilinger, one of the winners of the 2022 Nobel Prize in Physics. But there was one crucial difference: Zeilinger had been working with lasers and mirrors, rather than microwaves and superconductors.

For that reason, “we had to adapt the concept to the different experimental tools available for superconducting devices,” study co-author Gheorghe Sorin Paraoanu explained in a statement. Instead of light particles, the team used specially modified transmons – a type of superconducting qubit designed back in 2007 – to detect the presence of the microwave pulses.

“[We] had to change the standard interaction-free protocol in a crucial way: we added another layer of ‘quantumness’ by using a higher energy level of the transmon,” Paraoanu said. “Then, we used the quantum coherence of the resulting three-level system as a resource.”

“Quantum coherence” refers to that particular property that makes quantum mechanics so confusing. It’s the Schrödinger’s Cat paradox: the ability for objects to occupy two different states at the same time – even though under classical physics rules, that should be impossible. The quantum world, however, has no such problems with superpositions – and the team were able not just to work with this effect, but use it to their advantage.

The experiment was a success – and theoretical models confirmed their results. “We also demonstrated that even very low-power microwave pulses can be detected efficiently using our protocol,” added Shruti Dogra, fellow co-author of the paper.

All of which might leave you thinking, well, that’s cool, but it’s a bit niche, isn’t it? Here’s the kicker, though: this result has applications that range far wider than just a cute little demonstration of quantum weirdness.

“In quantum computing, our method could be applied for diagnosing microwave-photon states in certain memory elements,” Paraoanu pointed out. “This can be regarded as a highly efficient way of extracting information without disturbing the functioning of the quantum processor.”

Meanwhile, the team is already looking at further implications of their findings: applications such as counterfactual communication – that is, communication between two parties in which no physical particles are transferred – and counterfactual quantum computing, where the computations can yield results without the computer itself ever being run.

If that sounds bizarre or nonsensical to you, well, you’re not wrong. But in the quantum world, those kinds of mind-boggling concepts are really just a standard Thursday.

The study is published in the journal Nature Communications.

Source

It's almost hard to believe, but we're already nearing the end of yet another year, and 2022 was certainly been an eventful one for Microsoft. As Windows 11 made its way to more devices around the world, bringing a fresh new look and experience to users, Microsoft had a lot more to announce. Whether that's hardware in the form of new Surface devices, big Windows 11 updates, or the biggest acquisition of a gaming company ever, Microsoft delivered a lot — and fell short in a few ways, too.

With the year coming to an end, it's a good time to look back at all the major events and announcements Microsoft presented us with as we prepare for the year ahead.

The $70 billion Activision purchase

Microsoft started the year off with one of the biggest bombshells you could possibly expect. We were only in mid-January when Microsoft announced its bid to acquire video game developer and publisher Activision Blizzard for a whopping $68.7 billion.

This huge announcement represents the biggest acquisition ever in the gaming industry, a record that Microsoft itself had broken not that long ago with its acquisition of ZeniMax Media. It's also the biggest acquisition ever for Microsoft as a whole by a landslide (second place goes to GitHub, which Microsoft purchased for $26.2 billion back in 2016).

If the purchase is finalized, some of gaming's most iconic franchises will fall under Microsoft's wing, including the Call of Duty series, Diablo, Overwatch, and the mobile games developed and published by King, which includes Candy Crush Saga.

Microsoft wants to expand into mobile gaming, which is an increasingly large portion of the video game industry, and this acquisition is also meant to help with that, in addition to bolstering its Xbox platform. Interestingly enough, the deal could also benefit Nintendo fans, with Microsoft promising to bring the Call of Duty franchise to both Sony and Nintendo platforms for the next 10 years.

The deal to acquire Activision Blizzard has faced some legal challenges, and it's being reviewed by regulators in the U.K. and the U.S. Most recently, the Federal Trade Commission in the US announced it would file a suit to stop the acquisition from happening, though both Microsoft and Activision Blizzard remain confident that the deal will go through. It has yet to happen, but the announcement alone caused such a shockwave. And it was just the start of the year!

Android apps on Windows 11

The next big event for Microsoft was the launch of the Windows Subsystem for Android. Indeed, a feature that was initially promised for Windows 10 near the beginning of its existence is finally here with Windows 11: the ability to run Android apps on your Windows PC.

The Windows Subsystem for Android launched through a partnership with the Amazon Appstore, initially bringing around 1,000 apps to U.S. users. Since then, however, it has expanded significantly, with Microsoft now touting over 50,000 apps available through the Amazon Appstore on Windows 11, and you can get them in 31 countries. And if the apps on the Amazon Appstore aren't enough, more savvy users can sideload almost any app they want so long as Google services aren't required.

Microsoft has kept delivering improvements to the Android experience throughout the year, too. Controls for touch-based games are better, camera quality has improved, and even Android 13 is on the way soon. It's easier than ever to play your Android games on your PC or to develop and test Android apps on Windows.

Windows 11 version 22H2, or the 2022 Update

That's not all the news for Windows 11 users. After the big launch in October 2021, Microsoft began working on the next big Windows 11 update, officially called the Windows 11 2022 Update, or Windows 11 version 22H2. As the first major update for Windows 11, it packs many changes, with more being rolled out with smaller subsequent updates.

One of the most anticipated changes with this update was the addition of tabs in File Explorer, which Microsoft first announced in April after experimenting with the idea for a few years with Windows 10. Tabs make File Explorer work much like a web browser, so it becomes that much easier to manage your files without having to open multiple windows and manage them all. This is one of the features that rolled out with a smaller update in October, but it's a huge one all the same.

And there's more. The Start menu now has folders, you can change the size of the pinned and recommended areas, and the Task Manager has been completely redesigned. Microsoft also introduced a new Photos app shortly after, with a complete redesign and integration with iCloud Photos.

Microsoft even made some huge improvements to tablet use. Now, you can use swipe gestures to open or close the Start menu and navigate it. Similarly, the notifications and Quick Settings panels can be opened with swipe gestures, too. And when you're using full-screen apps, there's a new grabber, which helps prevent accidental swipes. Windows 11 tablets are just a lot better now.

The year when Arm devices got taken seriously

While it's not a single event, 2022 is also the year when Microsoft finally realized the importance of Arm-based devices, no doubt thanks to the booming success of Apple Silicon Macs. During its Build event, Microsoft announced its plans to offer a full developer toolchain for Arm users, including native support for Visual Studio, Windows Terminal, and more. And, of course, with most Android apps being designed for Arm, they should actually run better on Arm devices, too.

Part of this push for Arm is Project Volterra, officially known as the Windows Dev Kit 2023. This is an Arm-powered development machine meant to use all the developer tools that now support Arm. It makes it easier than ever to develop and test your Windows apps designed for Arm using a single device, hopefully resulting in many more Arm apps making their way to users in the near future. Microsoft itself updated many built-in Windows 11 apps to support Arm natively, like the Camera app, Notepad, and the Microsoft Store.

And of course, 2022 is the year we finally got Windows PCs with a much faster Arm processor, the Qualcomm Snapdragon 8cx Gen 3. Not many devices used it, but in addition to Project Volterra, we had the Lenovo ThinkPad 13s and Microsoft's own Surface Pro 9 with 5G, which are finally starting to deliver on everything Microsoft promised with Arm support. The best is yet to come, but 2022 certainly feels like a turning point for Windows on Arm devices.

Surface Laptop Go 2: A very welcome upgrade

Speaking of devices, it's time to talk about Microsoft's Surface products for 2022, which were generally underwhelming but featured good upgrades nonetheless. First, we got the Surface Laptop Go 2 in June, and it's a much-needed refresh for Microsoft's budget laptop.

This model didn't really change a lot from its predecessor, but the changes it does have are very welcome. Most notably, we've gone from an Intel Core i5-1035G1 to a Core i5-1135G7, which is a huge upgrade. Intel's 11th-generation processors deliver some big performance improvements on the CPU side, but even bigger improvements on the GPU side with Intel Iris Xe graphics. The original model was fine for school use, but the new one is legitimately good for the price.

This model also came with a slightly improved 720p webcam and a new color, Sage, which we would see again later in the year with the Surface Laptop 5. It's not rocking the world, but the Surface Laptop Go 2 is a "delightful entry-level laptop," as we said in our review.

Surface Laptop 5: A boring but necessary refresh

Towards the end of the year, Microsoft also delivered its annual fall hardware event, and with it came the Surface Laptop 5. This is another device that does the bare minimum to be a worthy upgrade, but sometimes the bare minimum is just enough.

There is one major change with the Surface Laptop 5 compared to its predecessor, and it's in the specs. The previous two models were offered in both AMD and Intel variants, but now, AMD has been removed from the lineup entirely, and all we get are 12th-generation Intel processors. They still offer a welcome performance upgrade, and Microsoft was smart enough to stick with 15W U-series processors instead of the more power-hungry P series, which many companies didn't do.

Not much else changed in the specs, but the one thing that needed to change did; the USB Type-C port now supports Thunderbolt 4, meaning you can connect a docking station, monitor, or an external GPU to the laptop to greatly increase its capabilities. Microsoft first brought Thunderbolt to its devices with the Surface Pro 9 and Laptop Studio, but this is the first mainstream laptop to feature support. Now, it's easier than ever to connect this laptop to a work setup at a desk and disconnect it when you're ready to leave.

However, a lot of things that should have been upgraded weren't. The design and display are the same we've been seeing for the last two generations (and, to some extent, since the first iteration), and the webcam is still only 720p.

Similar to the Surface Laptop Go 2, the 13-5-inch Surface Laptop 5 also came in a new Sage color, which replaced Ice Blue. This new color lacks the Alcantara wrist rest, so Platinum is now the only color option that offers it.

Surface Pro 9: The Surface Pro finally gets colors

Along with the Surface Laptop 5, Microsoft also introduced the Surface Pro 9. This is a big one. On the inside, it's mostly the same as before but with upgraded Intel 12th-generation processors. This delivers a notable performance boost, of course, which we won't complain about.

But the big news with this year's model was the introduction of color. Since its inception in 2012, the Surface Pro series has mostly only featured one color at a time. The first model was black, and then we got Platinum, which became the standard for years, until we finally got Matte Black/Graphite models again starting with the Surface Pro 6. But in 2022, we finally got real colors for the line. In addition to Platinum and Graphite, the Surface Pro 9 comes in a stunning Sapphire and an almost equally beautiful Forest colorway. These colors are actually much more vibrant than even the Surface Laptop series, and they look fantastic. And, of course, there are colored keyboards to match.

There are some downsides with this new model, however, most notably the removal of the headphone jack. The Surface Pro 9 has two Thunderbolt 4 ports and Surface Connect, but that's about it. Still, the colors alone make it a very exciting device.

Surface Pro 9 with 5G: A truly great Windows Arm tablet

In addition to the software improvements for Arm devices, Microsoft also delivered some big hardware improvements to its Arm-based tablets. The Surface Pro X has finally grown up and become the Surface Pro 9 with 5G, a sign from Microsoft that Arm is finally ready to be considered part of the main Surface Pro series.

The Surface Pro 9 with 5G is powered by the Microsoft SQ3 processor, a tweaked version of the Qualcomm Snapdragon 8cx Gen 3, and it brings a lot more performance to the table compared to its predecessors. It's also the first Surface device to support 5G instead of just LTE, so it should be more future-proof when it comes to connectivity.

But there's more. Microsoft put this hardware in the same chassis as the regular Surface Pro 9, which means it's slightly thicker now, and it also has a bigger battery. Thanks to that, the Surface Pro 9 with 5G is the first device that really delivers on the promise of all-day battery life with Windows on Arm. That's a big factor in portability, and it's great to finally see Microsoft deliver on it. Paired with always-on connectivity, this is a great device for working on the go.

We reviewed the Surface Pro 9 and found it to be the best Windows on Arm has ever been. It gives us hope for the future.

Surface Studio 2 Plus: The 'why?' moment of the year

Alongside all of these great devices, Microsoft introduced one of its oddest products in a long time, the Surface Studio 2 Plus. This is an upgrade to the four-year-old Surface Studio 2, and given that age, you might think it's a big one. In reality, this was a very underwhelming announcement with some questionable decisions.

For one thing, the Surface Studio 2 Plus barely changed the design of the product. The original Surface Studio 2 still looks very modern, but it has large bezels that some would consider unsightly today. That being said, Microsoft did make changes to the ports, adding three Thunderbolt 4 ports to the back to replace some of the Type-A ports in the Surface Studio 2.

The specs were also upgraded, but not as much as you might think. We get 35W 11th-generation processors, which are outdated and have a lower power rating than the ones on the previous model. Microsoft also upgraded to an Nvidia GeForce RTX 3060 Laptop GPU, which delivers significantly better graphics performance. Performance is improved, but it's still behind most premium all-in-one PCs in 2022.

What's truly baffling about all this is that without major design changes and outdated specs, Microsoft is still charging $4,500 for this device if you want the full experience. There's only one configuration available this time, so you don't have the option to spend less. The target audience for this product seems to be minuscule, and we have to wonder why Microsoft even bothered releasing it.

Microsoft Designer makes design easier than ever

Another one of the major announcements Microsoft made towards the end of the year was Microsoft Designer. This new design tool leverages the power of AI to make it easier than ever to start designing appealing images to share anywhere.

With Designer, social media managers can start with a simple text-based prompt to create the image they want. Using AI, Microsoft can instantly create images featuring the elements you want, like a picture of a cake or a scenic view. Designer also lets you add text or your own images to be used in the final design, creating richer and more visually appealing designs. Microsoft Designer is still in preview, but it should launch fully in 2023.

Alongside this, there's also Microsoft Create, which is a portal for all of Microsoft's design apps. Here, you can find templates and design ideas for all kinds of Microsoft apps, including PowerPoint, Clipchamp, Forms, and eventually, Designer.

Microsoft also launched the Image Generator in Bing, which also uses AI (specifically, DALL-E 2) to create images based on text input from users. It's all about making things as easy as possible for content creators and social media managers, and Microsoft's efforts in this space grew significantly in 2022.

All in all, Microsoft had quite a busy year in 2022, and while it's not the most exciting in every way, there was still some huge news. The Activision Blizzard purchase is still one of the biggest stories in the tech world right now, and we'll probably be hearing more about it well into 2023. We also saw real efforts to make Windows on Arm relevant, which was long overdue. And while some of Microsoft's devices were somewhat boring and predictable, they were still welcome upgrades. The exception, of course, is the Surface Studio 2 Plus, whose existence is nothing short of questionable.

Still, there were some very exciting news this year, and it's fair to say we'll be seeing the results of some of these announcements in 2023 and years to come.

Source

While many people only recently upgraded to Wi-Fi 6, and some may be considering a jump to Wi-Fi 6E, their successor is already on the horizon. Wi-Fi 7 is the next significant advance, and, just like its predecessors, the new standard promises faster connections, lower latency, and the ability to gracefully manage more connections than ever before.

If you are looking to improve your Wi-Fi today, Wi-Fi 7 is not the answer, because support will be scarce for some time to come. But early adopters will be excited by the news that the first batch of Wi-Fi 7 routers is set to land in early 2023. Realistically, it will be a good while before most of us should consider switching. If you need to upgrade pronto, first consider delving into how to buy a router before you check out our picks of the best Wi-Fi routers and the best mesh Wi-Fi systems.

Updated November 2022: We adjusted the timeline as Wi-Fi 7 draws closer and added news about TP-Link's Wi-Fi 7 router lineup.

What is Wi-Fi 7?

The 7th generation of Wi-Fi promises major improvements over Wi-Fi 6 and 6E and could offer speeds up to four times faster. It also includes clever advances to reduce latency, increase capacity, and boost stability and efficiency.

Wi-Fi 7 is IEEE 802.11be in the old naming convention, where Wi-Fi 6 was IEEE 802.11ax, and Wi-Fi 5 was IEEE 802.11ac. Like previous standards, Wi-Fi 7 will be backward compatible. But to take advantage of the new features and improved performance it promises, you will need to upgrade your devices. That means buying new routers and access points, not to mention new smartphones, laptops, TVs, and so on.

What Benefits Will Wi-Fi 7 Bring?

Wi-Fi 7 will be faster, support more connections, and be more adaptive to maintain reliable low latency performance.

These benefits will help deliver high-quality video and better cloud gaming, and they will serve AR and VR applications that require high throughput and low latency. Wi-Fi 7 also tackles congestion and interference, bringing tangible benefits to areas with densely packed devices or neighboring networks that overlap. The latter is most significant for the enterprise and larger venues.

How Does Wi-Fi 7 Compare to Wi-Fi 6E?

You may, understandably, wonder what distinguishes Wi-Fi 7 from Wi-Fi 6E, which broadly promises the same advantages over previous standards by opening up the 6-GHz band. Especially since Wi-Fi 7 will use the same three 2.4-GHz, 5 -GHz, and 6-GHz bands. Here are some notable upgrades.

Wider Channels

Each band is broken into channels. The 2.4-GHz band comprises 11 channels of 20 megahertz (MHz) each. The 5-GHz band has 45 channels, but instead of being limited to a width of 20 MHz, they can combine to create 40-MHz or 80-MHz channels. The 6-GHz band supports 60 channels, and with Wi-Fi 6E they can be as wide as 160 MHz. Wi-Fi 7 supports channels that are up to 320 MHz wide. The wider the channel, the more data it can transmit.

A simple analogy is to imagine how much traffic a single-lane road can handle compared to a three-lane highway or a six-lane superhighway.

Higher QAM

Quadrature Amplitude Modulation (QAM) is a method to transmit and receive data in radio-frequency waves. The higher it is, the more information you can pack in. Wi-Fi 7 supports 4K-QAM, Wi-Fi 6 supported 1024-QAM, and Wi-Fi 5 was limited to 256-QAM.

The potential benefits are complicated by signal strength, background noise, and interference. Consequently, as QAM goes up, the range drops, and you need a stronger signal. So, the jump to 1024-QAM in Wi-Fi 6 offered around a 25 percent data rate increase over Wi-Fi 5. The leap to 4K-QAM in Wi-Fi 7 translates to a 20 percent increase in peak performance.

Multi-Link Operation

Perhaps the most exciting advance in Wi-Fi 7 is Multi-Link Operation (MLO). Every previous Wi-Fi standard establishes a connection between two devices on a single band. Even a tri-band Wi-Fi 6E router connects two devices on a single band on a fixed channel (the router decides whether to connect on the 2.4-GHz, 5-GHz, or 6-GHz band).

MLO can combine several frequencies across bands into a single connection. A Wi-Fi 7 router can connect to a Wi-Fi 7 device across two or more channels in different bands simultaneously. MLO potentially enables wider channels capable of transmitting more data—going back to our highway analogy, you can send traffic on the highway and the superhighway at once.

Speed isn’t always the priority, but MLO also allows for more efficient performance. A Wi-Fi 7 router can take congestion and other interference into account and transmit on the best channel to bypass it, switching to maintain a stable connection and low latency. MLO may also help mitigate the relatively short range of the 6-GHz band.

Wi-Fi 7 also enhances existing technologies like OFDMA, MU-MIMO, and TWT, which we discuss in our Wi-Fi 6E explainer.

When Will Wi-Fi 7 Be Available?

The first Wi-Fi 7 devices and routers will start rolling out early in 2023. While you won’t feel the benefit of a Wi-Fi 7 router until you have Wi-Fi 7 devices, because the new Wi-Fi 7 routers are backward compatible, they may make sense for people currently considering an expensive Wi-Fi 6E system, as they will handle 6E connections but also afford some future-proofing.

Photograph: TP-Link

TP-Link has announced a whole Wi-Fi 7 router product line, including the flagship quad-band Archer BE900 ($700), which sports a distinctive X-shaped design with a customizable LED grid and a touchscreen. It boasts two 10 Gbps ports, four 2.5 Gbps ports, and a combined speed across bands of up to 24 Gbps (the 6-GHz band goes up to 11 Gbps, and there are two 5-GHz bands and a 2.4-GHz band). The company also unveiled the Archer GE800 Gaming router and three Deco mesh systems (though the headlining Deco BE95 will cost you a whopping $1,200 for a 2-pack).

On the device side, Qualcomm has its Wi-Fi 7 chipset and the Networking Pro Series platform, which can deliver up to 33 Gbps of quad-band connectivity over 16 streams. Partners are already working on integrating its technologies into devices. Broadcom and MediaTek have also announced Wi-Fi 7 technologies. More vendors will undoubtedly announce their plans in the coming months as the new standard is firmed up and officially designated by the Wi-Fi Alliance.

Even when Wi-Fi 7 does arrive, it’s not going to sweep Wi-Fi 6 away. The two will likely coexist as complementary technologies for many years to come.

What is Wi-Fi 7? Everything You Need to Know

(May require free registration to view)

According to the press release, TSMC estimates that 3nm technology will create end products with a market value of $1.5 trillion within five years of mass production. TSMC has a reputation for producing high-quality chips using advanced manufacturing processes, and it serves a wide range of customers in the tech industry, including Apple, Qualcomm, and Nvidia.

In addition to the success of 3nm technology, TSMC highlighted its expansion of 3nm capacity in Taiwan and Arizona. The company's Fab 18 facility in the Southern Taiwan Science Park will serve as its "GIGAFAB" facility producing 5nm and 3nm process technology, while the Arizona site will also have 3nm capacity. These expansions demonstrate TSMC's commitment to investing in advanced manufacturing capabilities and meeting the demand for 3nm technology.

TSMC is also making preparations for its 2nm fabs, which will be located in the Hsinchu and Central Taiwan Science Parks. These fabs will consist of six phases and are proceeding as planned. Previously, the company had also announced its plans for manufacturing 4nm chips at its Arizona fab when it opens in 2024 due to increasing demand from Apple.

TSMC mentioned how 3nm is the most advanced semiconductor technology in terms of power, performance, and area (PPA) and offers a 1.6X logic density gain and 30-35% power reduction compared to its 5nm process. It also has advanced transistor technology and is expected to support the development of new applications in areas such as artificial intelligence, 5G, and high-performance computing.

TSMC says that it is committed to green manufacturing and has implemented various measures to reduce its environmental impact and promote sustainability. The company's Fab 18 facility in the Southern Taiwan Science Park follows Taiwan's EEWH and the U.S. LEED green building certification standards. The facility will also use water resources from the TSMC STSP Reclaimed Water Plant to gradually reach the company's target of using 60% reclaimed water by 2030.

Once mass production begins at Fab 18, the facility will use 20% renewable energy to eventually reach the sustainability goal of 100% renewable energy and zero emissions by 2050.

TSMC expands 3nm capacity in Taiwan and Arizona, prepares for 2nm fabs

Amazon begins drone deliveries in California and Texas

In August, when AMD announced Zen 4 based Ryzen 7000 series processors, it released just 4 CPUs in the lineup. The AMD Ryzen 9 7950X, Ryzen 9 7900X, Ryzen 7 7700X and the Ryzen 5 7600X.

Couple of months later, a famous Twitter based leaker revealed that AMD is planning more processors, this time the processors would be non-X ones. As per the latest leaked slides, that leaker seems to have got it spot on.

AMD Ryzen 7600, 7700 & 7900 Slides Leaked

VideoCardz seems to have gotten hold onto leaked slides which confirm AMD Ryzen 7600, 7700 & 7900 processors. It also reveals that these processors are having a release date of 10th January 2023.

The AMD Ryzen 9 7900 is going to be a 12 core, 24 thread processor just like the 7900X. However, 7900 will have a slightly lesser boost clock of 5.4GHz vs 5.6GHz of 7900X. The AMD Ryzen 9 7900 will cost $429, compared to the original MSRP of 7900X, which is $549. Making it a whole $120 cheaper.

The AMD Ryzen 7 7000 is going to be an 8 core, 16 thread processor like the 7700X. The boost clock here is very slightly lesser at 5.3GHz vs 5.4GHz of 7700X. It’s going to be priced at $329. Which is $70 lesser than the original MSRP of 7700X.

Now comes the cheapest processor in the lineup so far. The AMD Ryzen 5 7600 is going to be a 6 core, 12 thread processor, again, same as 7600X. The boost clock of 7600 is going to be 5.1GHz compared to the 5.3GHz of 7600X. The AMD Ryzen 7600 is going to cost $229 compared to the 7600X which had an original MSRP of $299. Again, making it $70 cheaper.

However, there’s one thing that sets all these three processors apart from their X variants. All the above three processors come with just 65W TDP. The 7900X had an TDP of 170W, while the 7700X and 7600X had a TDP of 105W. This means that these processors will have a hard power usage limit, which means the performance might be slightly limited too. We currently don’t know whether these CPUs will be available to the public or will be made available only for PC making companies.

AMD Ryzen 7900 Performance Revealed

VideoCardz got hold onto another slide which compares the performance of AMD Ryzen 7900 with the Ryzen 5900X processor. In the twelve games tested, the Ryzen 7900 is about 20% faster than the Ryzen 5900X, which is quite good.

Nothing unexpected, but we are talking about a newer gen $429 beating a $549 previous generation processor in performance. So it’s something nice to have.

Outcome Of These New Processors

AMD Ryzen 7000 series are not selling much. Reddit based user, AverageEnjoyer2023, has posted that as per German retailer Mindfactory, the difference between Ryzen 7600X and 5600X sales is huge.

Every day, sixteen 7600X CPUs are being sold. Compared to that, they are selling ninety 5600X processors per day. This means a massive 6x difference in sales.

The reason is simple. While the new Ryzen 7000 processors aren’t bad, nor are they too expensive. It’s the whole platform that makes these new processors unattractive. The Ryzen 7000 processors need new motherboards, which are very expensive currently because they need to support both new and expensive AM5 CPU slot and also DDR5 RAM. DDR5 RAM sticks themselves are expensive too.

Not to forget, some motherboards come with an PCIe 5.0 slot, whereas no graphics card currently uses it. Even the SSDs based on PCIe 5.0 currently are rare or don’t make a full use of it due to NAND chips (used in SSDs) required for such speeds have started mass production only now.

So basically, making a Ryzen 5000 series computer is far cheaper than the Ryzen 7000 series computer right now. So unless the above issues are resolved, the non-X Ryzen 7000 processors don’t give much to the users.

But that’s not it. The non-X Ryzen 7000 processors don’t offer much in terms of price cut. This is because after recent price cuts by AMD on the X processors, the difference between X and non-X Ryzen processors is only going to be $10-20. Which is not at all impressive.

Whatever it will be. AMD has long term considerations with its new AM5 platform. Whether it brings them any rewards, is to be seen.

AMD To Release Ryzen 7600, 7700 & 7900 CPUs On 10th January

From left to right and largest to smallest: GeForce RTX 4080 (which is the same physical size as the RTX 4090), Radeon RX 7900 XTX, and Radeon RX 7900 XT.

Andrew Cunningham

 

In 2021, the biggest story about GPUs was that you mostly just couldn't buy them, not without paying scalper-inflated prices on eBay or learning to navigate a maze of stock-tracking websites or Discords.

 

The good news is that the stock situation improved a lot in 2022. A cryptocurrency crash and a falloff in PC sales reduced the demand for GPUs, which in turn made them less profitable for scalpers, which in turn improved the stock situation. It's currently possible to visit an online store and buy many GPUs for an amount that at least gets kind-of-sort-of close to their original list price.

 

We also saw lots of new GPU launches in 2022. The year started off less-than-great with the launch of 1080p-focused, price-inflated cards like Nvidia's RTX 3050 and AMD's inspiringly mediocre RX 6500 XT. But by the end of the year, we received Nvidia's hugely expensive but hugely powerful RTX 4090 and RTX 4080 cards, AMD's less-monstrous but still competitive RX 7900 series, and Intel's flawed but price-conscious Arc A770 and A750 cards.

 

The bad news is that the aftereffects of the GPU shortage still linger, mainly in the form of inflated prices. We can hope that these come down in 2023, but so far, there is little sign of that happening.

Budget GPUs are in a sad state

You can still find GPUs at and under $200 if you're looking for basic, better-than-integrated performance in older and lower-end games that you'll mostly run at or below 1080p.

 

But performance in this category has moved very little over the last three or four years. Nvidia seems content to serve this low-end slice of the gaming market with the same GeForce GTX 1650 GPU it introduced in 2019, a card that continues to stubbornly hover in the $150-to-$200 price window despite its age. AMD and Intel have both released new cards for the sub-$200 market in the last year, and those cards can sometimes beat the GTX 1650's performance. But these cards are also flawed in some hard-to-ignore ways.

 

AMD's RX 6500 XT was originally a laptop GPU that was adapted for desktops, and as a result, it supports fewer displays than other GPUs in the RX 6000 series, it's missing hardware video encoding support, and its performance in older PCI Express 3.0-capable PCs is poor because it only provides four lanes of PCIe bandwidth in the first place. Intel's Arc A380 has great video encoding support (including for the AV1 video codec), but like other Arc cards, its drivers are rough around the edges, and performance in older games can be spotty.

 

If superior GPUs like Nvidia's RTX 3050 series and AMD's RX 6600 series drop down into the $200-and-under range soon, we'll be feeling a lot better about the state of budget GPUs (the RX 6600 is getting awful close, with prices falling into the $220-250 range for some models depending on the sale). Those GPUs still come with compromises, like mediocre ray-tracing performance and a more hit-or-miss experience at resolutions above 1080p, but it'll be nice to put the GTX 1650 firmly in the rear-view.

For the best GPUs, $1,000 and up is the new normal

Shifting to the other extreme of the market, it used to be that GPUs with four-digit price tags were mostly ignorable by normal people. Halo products like Nvidia's Titan GPUs performed well, sure, but why pay all that money when cheaper xx80 and xx70-series GPUs could give you a large percentage of that performance for a small percentage of the price?

 

The story of this generation's midrange cards still has yet to be written, but so far, the high-end cards in the RTX 4000 series and the RX 7000 series have been listed for much more than their predecessors. The $1,200 RTX 4080 is a huge jump from the $700 Nvidia originally advertised for the RTX 3080 and 2080. If the newly rebranded 4070 Ti launches at $900 as is currently rumored (and was originally planned, back when it was called the "RTX 4080 12GB"), that will be a substantial hike over the $500-to-$600 launch prices of cards like the RTX 2070, RTX 3070, and RTX 3070 Ti.

 

All of these prices are made much worse by the fact that you can't find RTX 4080 or 4090 cards anywhere close to their launch MSRPs right now.

 

Things look marginally less exorbitant on AMD's side, where the top-tier RX 7900 XTX launched for the same $999 price as last-generation's top-tier RX 6900 XT did. But where the 6900 XT was accompanied by an RX 6800 XT that provided most of the performance for $649, the 7900 XTX's little sibling is an $899 RX 7900 XT that actually gives you less performance-per-dollar than the XTX. Both cards still undercut Nvidia's pricing for the RTX 4000 series, but that's more about how expensive the 4090 and 4080 are than it is about the value AMD is providing.

 

Nvidia CEO Jensen Huang said that these higher prices are here to stay, courtesy of the increased costs associated with designing and manufacturing these GPUs. Obviously, Huang is not an impartial observer here—he is materially invested in keeping the prices of his GPUs high, especially as Nvidia's finances have suffered. But there's some truth to what he's saying: cutting-edge manufacturing processes are expensive, Nvidia is fighting AMD, Apple, and all kinds of other chip designers for TSMC's production capacity, and gigantic monolithic chips like the RTX 4000 GPUs are going to have lower yields than smaller, less-complex processor dies.

2022 in GPUs: The shortage ends, but higher prices seem here to stay

With Brainomix e-Stroke, doctors are given real-time decision support from the AI, which looks at brain scans and attempts to suggest the right treatment while considering how long it has been since the stroke happened. The software also allows stroke specialists to access scans and images remotely so they can assist remotely too.

The AI system is deployed at sites across 11 stroke networks in England. It has been used to detect over 4,500 large vessel occlusions in stroke patients. Brainomix e-Stroke is particularly helpful for detecting LVOs because they require time-sensitive diagnoses and the sooner they’re treated, the more likely a patient is to recover.

“Every minute saved during the initial hospital assessment of people with stroke-like symptoms can dramatically improve a patient’s chance of leaving hospital in good health,” NHS England Director of Transformation Dr Timothy Ferris said about the software. “The NHS is harnessing the potential that AI has to support expert staff in delivering life-changing care for patients with a range of needs, and through the AI in Health and Care awards we are testing, evaluating and supporting the most promising technologies which could transform the way we deliver care.”

So far, government funding for the Brainomix e-Stroke system has helped treat 111,000 suspected stroke patients around England. It’s just the latest example of how artificial intelligence is improving people’s health.

Source

Feature As word of students using AI to automatically complete essays continues to spread, some lecturers are beginning to rethink how they should teach their pupils to write.

Writing is a difficult task to do well. The best novelists and poets write furiously, dedicating their lives to mastering their craft. The creative process of stringing together words to communicate thoughts is often viewed as something complex, mysterious, and unmistakably human. No wonder people are fascinated by machines that can write too.

Unlike humans, language models don't procrastinate and create content instantly with a little guidance. All you need to do is type a short description, or prompt, instructing the model on what it to do and it'll generate a text output in seconds. So it should come as no surprise students are now beginning use these tools to complete school work.

Students are the perfect users: They need to write often, in large volumes, and are internet savvy. There are many AI-writing products to choose from that are easy to use and pretty cheap too. All of them lure new users with free trials, promising to make them better writers.

Monthly subscriptions for the most popular platform, Jasper, costs $40 per month to generate 35,000 words. Others, like Writesonic or Sudowrite, are cheaper at $10 per month for 30,000 words. Students who think they can use these products and get away with doing zero work, however, will probably be disappointed.

And then there's ChatGPT...

Although AI can generate text with perfect spelling, great grammar and syntax, the content often isn't that good beyond a few paragraphs. The writing becomes less coherent over time with no logical train of thought to follow. Language models fail to get their facts right – meaning quotes, dates, and ideas are likely false. Students will have to inspect the writing closely and correct mistakes for their work to be convincing.

Prof: AI-assisted essays 'not good'

Scott Graham, associate professor at the Department of Rhetoric & Writing at the University of Texas at Austin, tasked his pupils with writing a 2,200-word essay about a campus-wide issue using AI. Students were free to lightly edit and format their work with the only rule being that most of the essay had to be automatically generated by software.

In an opinion article on Inside Higher Ed, Graham said the AI-assisted essays were "not good," noting that the best of the bunch would have earned a C or C-minus grade. To score higher, students would have had to rewrite more of the essay using their own words to improve it, or craft increasingly narrower and specific prompts to get back more useful content.

"You're not going to be able to push a button or submit a short prompt and generate a ready-to-go essay," he told The Register.

The limits of machine-written text forces humans to carefully read and edit copy. Some people may consider using these tools as cheating, but Graham believes they can help people get better at writing.

Don't waste all your effort on the first draft....

"I think if students can do well with AI writing, it's not actually all that different from them doing well with their own writing. The main skills I teach and assess mostly happen after the initial drafting," he said.

"I think that's where people become really talented writers; it's in the revision and the editing process. So I'm optimistic about [AI] because I think that it will provide a framework for us to be able to teach that revision and editing better.

"Some students have a lot of trouble sometimes generating that first draft. If all the effort goes into getting them to generate that first draft, and then they hit the deadline, that's what they will submit. They don't get a chance to revise, they don't get a chance to edit. If we can use those systems to speed write the first draft, it might really be helpful," he opined.

Whether students can use these tools to get away with doing less work will depend on the assignment. A biochemistry student claimed on Reddit they got an A when they used an AI model to write "five good and bad things about biotech" in an assignment, Vice reported.

AI is more likely to excel at producing simple, generic text across common templates or styles.

Listicles, informal blog posts, or news articles will be easier to imitate than niche academic papers or literary masterpieces. Teachers will need to be thoughtful about the essay questions they set and make sure students' knowledge are really being tested, if they don't want them to cut corners.

Ask a silly question, you'll get a silly answer

"I do think it's important for us to start thinking about the ways that [AI] is changing writing and how we respond to that in our assignments -- that includes some collaboration with AI," Annette Vee, associate professor of English and director of the Composition Program at the University of Pittsburgh, told us.

"The onus now is on writing teachers to figure out how to get to the same kinds of goals that we've always had about using writing to learn. That includes students engaging with ideas, teaching them how to formulate thoughts, how to communicate clearly or creatively. I think all of those things can be done with AI systems, but they'll be done differently."

The line between using AI as a collaborative tool or a way to cheat, however, is blurry. None of the academics teaching writing who spoke to The Register thought students should be banned from using AI software. "Writing is fundamentally shaped by technology," Vee said.

"Students use spell check and grammar check. If I got a paper where a student didn't use these, it stands out. But it used to be, 50 years ago, writing teachers would complain that students didn't know how to spell so they would teach spelling. Now they don't."

Most teachers, however, told us they would support regulating the use of AI-writing software in education. Anna Mills, who teaches students how to write at a community college in the Bay Area, is part of a small group of academics beginning to rally teachers and professional organizations like the Modern Language Association into thinking about introducing new academic rules.

Critical thinking skills

Mills said she could see why students might be tempted to use AI to write their essays, and simply asking teachers to come up with more compelling assessments is not a convincing solution.

"We need policies. These tools are already pretty good now, and they're only going to get better. We need clear guidance on what's acceptable use and what's not. Where is the line between using it to automatically generate email responses and something that violates academic integrity?" she asked The Register.

"Writing is just not outputs. Writing and revising is a process that develops our thinking. If you skip that, you're going to be skipping that practice which students need.

"It's too tempting to use it as a crutch, skip the thinking, and skip the frustrating moments of writing. Some of that is part of the process of going deeper and wrestling with ideas. There is a risk of learning loss if students become dependent and don't develop the writing skills they need."

Mills was particularly concerned about AI reducing the need for people to think for themselves, considering language models carry forward biases in their training data. "Companies have decided what to feed it and we don't know. Now, they are being used to generate all sorts of things from novels to academic papers, and they could influence our thoughts or even modify them. That is an immense power, and it's very dangerous."

Lauren Goodlad, professor of English and Comparative Literature at Rutgers University, agreed. If they parrot what AI comes up with, students may end up more likely to associate Muslims with terrorism or mention conspiracy theories, for example.

Computers are alredy interfering and changing the ways we write. Goodlad referred to one incident when Gmail suggested she change the word "importunate" to "impatient" in an email she wrote.

"It's hard to teach students how to use their own writing as a way to develop their critical thinking and as a way to express knowledge. They very badly need the practice of articulating their thoughts in writing and machines can rob them of this. If people really do end up using these things all the way through school, if that were to happen it could be a real loss not just for the writing quality but for the thinking quality of a whole generation," she said.

Rules and regulation

Academic policies tackling AI-assisted writing will be difficult to implement. Opinions are divided on whether sentences generated by machines count as plagiarism or not. There is also the problem of being able to detect writing produced by these tools accurately. Some teachers are alarmed at AI's growing technical capabilities, whilst others believe its overhyped. Some are embracing the technology more than others.

Marc Watkins, lecturer, and Stephen Monroe, chair and assistant professor of writing and rhetoric, are working on building an AI writing pilot programme with the University of Mississippi's Academic Innovations Group. "As teachers, we are experimenting, not panicking," Monroe told The Register.

"We want to empower our students as writers and thinkers. AI will play a role… This is a time of exciting and frenzied development, but educators move more slowly and deliberately… AI will be able to assist writers at every stage, but students and teachers will need tools that are thoughtfully calibrated."

Teachers are getting together and beginning to think about these tools, Watkins added. "Before we have any policy about the use of language models, we need to have sustained conversations with students, faculty, and administration about what this technology means for teaching and learning."

"But academia doesn't move at the pace of Big Tech. We're taking our time and slowly exploring. I don't think faculty need to be frightened. It's possible that these tools will have a positive impact on student learning and advancing equity, so let's approach AI assistants cautiously, but with an open mind."

Regardless of what policies universities may decide to implement in the future, AI presents academia with an opportunity to improve education now. Teachers will need to adapt to the technology if they want to remain relevant, and incentivise students to learn and think on their own with or without assistance from computers. ®

Source

A newly leaked slide published by VideoCardz has spilled the beans on AMD's upcoming alleged non-X Ryzen 7000 processors. The company apparently plans to announce lower-TDP models on January 10, 2023, right around time for CES 2023, providing customers with more affordable variants of the current lineup.

AMD will lower the base and boost clocks to fit the new processors into a lower TDP envelope from 170W and 105W to only 65W while retaining the same number of cores and threads. Of course, the price will also go down as a tradeoff for a performance downgrade.

The upcoming Ryzen 9 7900, Ryzen 7 7700, and Ryzen 5 7600 models will receive $429, $329, and $229 price tags. Here is how the processors compare to their more expensive X variants:

Performance-wise, the leaked slide claims the Ryzen 9 7900 is about 19% faster on average than its counterpart from the Ryzen 5000 Series when paired with the AMD RX 6950XT.

Even with lower clocks and TDP, the upcoming non-X Ryzen 7000 processors should offer decent performance for lower prices. However, AMD's latest CPU lineup is currently on sale, with massive discounts reaching up to 29%.

Source: VideoCardz

AMD reportedly to launch three non-X Ryzen 7000 processors on January 10

In August 1955, a group of scientists made a funding request for US$13,500 to host a summer workshop at Dartmouth College, New Hampshire. The field they proposed to explore was artificial intelligence (AI).

While the funding request was humble, the conjecture of the researchers was not: "Every aspect of learning or any other feature of intelligence can, in principle, be so precisely described that a machine can be made to simulate it".

Since these humble beginnings, movies and media have romanticized AI or cast it as a villain. Yet, for most people, AI has remained as a point of discussion and not part of a conscious lived experience.

AI has arrived in our lives

Late last month, AI, in the form of ChatGPT, broke free from the sci-fi speculations and research labs and onto the desktops and phones of the general public.

It's what's known as a "generative AI" – suddenly, a cleverly worded prompt can produce an essay or put together a recipe and shopping list, or create a poem in the style of Elvis Presley.

While ChatGPT has been the most dramatic entrant in a year of generative AI success, similar systems have shown even wider potential to create new content, with text-to-image prompts used to create vibrant images that have even won art competitions.

AI may not yet have a living consciousness or a theory of mind popular in sci-fi movies and novels, but it is getting closer to at least disrupting what we think artificial intelligence systems can do.

Researchers working closely with these systems have swooned under the prospect of sentience, as in the case with Google's large language model (LLM) LaMDA. An LLM is a model that has been trained to process and generate natural language.

Generative AI has also produced worries about plagiarism, exploitation of original content used to create models, ethics of information manipulation and abuse of trust, and even "the end of programming".

At the center of all this is the question that has been growing in urgency since the Dartmouth summer workshop: Does AI differ from human intelligence?

What does 'AI' actually mean?

To qualify as AI, a system must exhibit some level of learning and adapting. For this reason, decision-making systems, automation, and statistics are not AI.

AI is broadly defined in two categories: artificial narrow intelligence (ANI) and artificial general intelligence (AGI). To date, AGI does not exist.

The key challenge for creating a general AI is to adequately model the world with all the entirety of knowledge, in a consistent and useful manner.

That's a massive undertaking, to say the least.

Most of what we know as AI today has narrow intelligence – where a particular system addresses a particular problem. Unlike human intelligence, such narrow AI intelligence is effective only in the area in which it has been trained: fraud detection, facial recognition, or social recommendations, for example.

AGI, however, would function as humans do. For now, the most notable example of trying to achieve this is the use of neural networks and "deep learning" trained on vast amounts of data.

Neural networks are inspired by the way human brains work. Unlike most machine learning models that run calculations on the training data, neural networks work by feeding each data point one by one through an interconnected network, each time adjusting the parameters.

As more and more data are fed through the network, the parameters stabilize; the final outcome is the "trained" neural network, which can then produce the desired output on new data – for example, recognizing whether an image contains a cat or a dog.

The significant leap forward in AI today is driven by technological improvements in the way we can train large neural networks, readjusting vast numbers of parameters in each run thanks to the capabilities of large cloud-computing infrastructures. For example, GPT-3 (the AI system that powers ChatGPT) is a large neural network with 175 billion parameters.

What does AI need to work?

AI needs three things to be successful.

First, it needs high-quality, unbiased data, and lots of it. Researchers building neural networks use the large data sets that have come about as society has digitized.

Co-Pilot, for augmenting human programmers, draws its data from billions of lines of code shared on GitHub. ChatGPT and other large language models use the billions of websites and text documents stored online.

Text-to-image tools, such as Stable Diffusion, DALLE-2, and Midjourney, use image-text pairs from data sets such as LAION-5B. AI models will continue to evolve in sophistication and impact as we digitize more of our lives and provide them with alternative data sources, such as simulated data or data from game settings like Minecraft.

AI also needs computational infrastructure for effective training. As computers become more powerful, models that now require intensive efforts and large-scale computing may, in the near future, be handled locally. Stable Diffusion, for example, can already be run on local computers rather than cloud environments.

The third need for AI is improved models and algorithms. Data-driven systems continue to make rapid progress in domain after domain once thought to be the territory of human cognition.

However, as the world around us constantly changes, AI systems need to be constantly retrained using new data. Without this crucial step, AI systems will produce answers that are factually incorrect or do not take into account new information that's emerged since they were trained.

Neural networks aren't the only approach to AI. Another prominent camp in artificial intelligence research is symbolic AI – instead of digesting huge data sets, it relies on rules and knowledge similar to the human process of forming internal symbolic representations of particular phenomena.

But the balance of power has heavily tilted toward data-driven approaches over the last decade, with the "founding fathers" of modern deep learning recently being awarded the Turing Prize, the equivalent of the Nobel Prize in computer science.

Data, computation, and algorithms form the foundation of the future of AI. All indicators are that rapid progress will be made in all three categories in the foreseeable future.

George Siemens, Co-Director, Professor, Centre for Change and Complexity in Learning, University of South Australia

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Source

In brief Sundar Pichai is apparently all in a pickle over OpenAI's ChatGPT engine, and is gearing up Google to meet the perceived threat.

According to an internal memo seen by the New York Times, Pichai has "upended the work of numerous groups inside the company to respond to the threat that ChatGPT," and is plucking staff from other divisions to meet the threat to the OpenAI's plans. It's reportedly considered a "Code Red" for the Chocolate Factory.

At issue is whether Google's core product, search, will be displaced by AI systems that can give more accurate research results, and that's a big if, for the moment at least.

"No company is invincible; all are vulnerable," said Margaret O'Mara, a professor at the University of Washington. "For companies that have become extraordinarily successful doing one market-defining thing, it is hard to have a second act with something entirely different."

The report suggests that Google will make a series of AI announcements on May to meet the growing threats to the search giant's business model. We'll see if these are functional products or just Google playing catch up.

Google has dominated the search market for 20 years, and anything that threatens that highly lucrative business - which makes up around 90 percent of Alphabet's profits - is something Sundar might well have reason to fear.

ArtStation cracks down on anti-AI art protests

The ongoing fight between human artists and ArtStation, the Epic Games-owned site that displays the images and, it's claimed, exploits the data for AI purposes, has stepped up a notch.

Last week many users of the site protested at the use of their uncredited images to train AI generation models for art. The fear is that ArtStation is allowing AI trainers to take legitimate human work and not only create art, but also potentially drive artists out of business. In response artists started posting "AI is theft" banners on their profile pages.

Now ArtStation has reportedly lowered the bar and is banning such subversive creations. "For site usability, we are moderating posts that violate our Terms of Service," it said on Twitter.

"We understand concerns about AI and its impact on the industry. We will share more about improvements to give users more control over what they see and how they use ArtStation in the near future."

In other words, such it up you creative types. This one is likely to play out for some time.

US senator shuts the door on AI as he walks out

The outgoing Senator Rob Portman (D-OH) has introduced the Facial Accountability, Clarity, and Efficiency In Technology Act (FACE IT) to Congress calling for much tighter controls on the US federal government using AI-powered facial recognition technology.

The statute would require the National Institute of Standards and Technology (NIST) to set minimum acceptable accuracy standards for facial recognition technology, allow citizens an opt-out from being identified solely by such systems. It also wants to ensure a human authority must give authority for such systems to be used.

"Facial recognition technology can be used to help protect our communities, but I am concerned about the potential for abuse," Portman, who leaves Congress in January, said.

"I'm proud to introduce the FACE IT Act because, given the civil liberty implications of the federal government's use of facial recognition technology, we must pass legislation to set rules for the use of this technology. We must make sure federal law enforcement and other agencies have the tools to do their jobs well, but it is vital that we set rules for those tools."

He also introduced the Stopping Unlawful Negative Machine Impacts through National Evaluation Act, which would "clarify that existing civil rights laws apply to decisions made by AI systems just as if those decisions were made by humans."

The proposed laws, which seemingly have little chance of making it onto the statute books given the fractious state of Congress, seems mostly about publicity and a possible future lobbying career than an attempt to fix solid policy in place.

Source

Tech and business journalists know what to do with Twitter.

Twitter is a for-profit company with headquarters in California. It has a CEO. It has investors and revenues and a valuation. The purpose of the company is to make money for its investors by growing the user base as large as possible and extracting as much cash as possible from that user base, either through advertising or subscriptions. Though the product it delivers is different, Twitter is pretty much the same thing as Snapchat and Facebook and Netflix, and Google.

Twitter fits easily into the standard tech journalism templates, which include:

•     “New, exciting tech company will revolutionize everything!”

•     “Older, established tech company releases new product!”

•     “Look how rich or weird this tech CEO is!”

•     “Tech company acquires another tech company!”

•     “Tech company is on the rise!”

•     “Tech company is in decline!”

•     And, when things go awry: “Tech founder faces criminal charges!”

Tech journalism isn’t special in this regard. Every genre of journalism has narrative grooves — ruts may be a better word — into which it easily slides.

A lot of what we see in tech journalism is either consumerist content or standard business journalism dressed up in a Steve Jobs turtleneck. The consumerist stuff shows up as weirdly breathless reviews of new products that are mostly the same as the old products or descriptions of updates to existing tech (“three exciting new features in iOS 16.2!”). The business-journalism stuff is all about stock prices and mergers and acquisitions. For some reason, people who would never read an article in the Wall Street Journal about Procter and Gamble acquiring a new shampoo company will devour content about Microsoft acquiring a tech startup they’ve never heard of.

Tech journalism serves these purposes pretty well. I have certainly read my share of reviews of tech products and I am far more familiar with the business activities of Elon Musk and Jeff Bezos than I really need to be. But what happens when something big happens in the tech world that doesn’t fit the standard narratives?

Values over profit

I was prompted to write this post when I ran across the following post on Mastodon by Annalee Newitz:
 

Gotta love tech journalists who describe Mastodon as “that impossible-to-use website.” First of all, it’s an app. C’mon. Second of all, aren’t these the same people who write breathless explainers about the wonder of cryptocurrencies, which are not only impossible to understand but literally built from bullshit?

Like Newitz, I’m an increasingly enthusiastic adopter of Mastodon, and, like them, I’ve been kind of confused by the press coverage around the platform. The media seems to be regarding Mastodon as a bizarre curiosity, something that the general public couldn’t possibly grasp. Sure, the guys with a Linux server in their basement might geek out on it, but this thing isn’t for the masses.

Maybe.

But maybe the problem is that tech journalists don’t know what to do with something like Mastodon. You see, Mastodon doesn’t fit the standard tech narratives.

Mastodon is not a company. It’s an open-source platform that isn’t owned by anybody. It has a founder, a German guy named Eugen Rochko, who seems to be the opposite of the standard megalomaniacal tech founder. He started Mastodon as a personal project, and he maintains it along with hundreds of other volunteers via Github. Nobody controls Mastodon. It’s decentralized on purpose — a person or institution can set up their own Mastodon server, which can communicate with other servers just like you can email somebody on Outlook from your Gmail account.

Perhaps most importantly, no one is profiting off of Mastodon. Unlike almost everything in the tech industry, it wasn’t started as a way to make its founders piles of cash.

What’s attractive about Mastodon isn’t the software (it’s not as slick as corporate social media but it’s still very good) — it’s the values of the platform. No one is trying to hack the attention of Mastodon users for profit, no one is bombarding us with ads. It’s just a community of people, communicating.

Mastodon’s values are antithetical to the way we’ve come to understand the purpose of technology and technology companies. We’ve been trained to favor ease of use and affordability over everything else. If a product is slick or free, then we are expected to sacrifice whatever’s necessary — usually our privacy or our attention span — to use it.

Tech journalism usually operates on and amplifies these assumptions; they’re baked into most of the coverage of the technology world.

But what if people decided that the values of a platform were the most important thing about it? What if people liked the feeling of not being manipulated by algorithms? What if people thought that not feeling like garbage every time they log off of social media was more important than having a really slick app? What if one of our big social networks was crowdfunded and run mostly by volunteers?

Right now, using Mastodon feels way better than using Twitter did. People there are more engaged and less cynical. There’s more communication and less broadcasting for likes and follows. It feels healthier. This may change, of course — I worry that if Mastodon gets big enough, for-profit companies may stomp in somehow. But at the moment, it feels refreshingly different.

So far, tech journalists have approached Mastodon like it’s another startup. By trying to force Mastodon into their usual narratives, they’re missing that what’s really innovative about Mastodon isn’t its technology — it’s the platform’s values.

Source