These kinds of issues are nothing but frustrating, as they damage your fun and annoy you in many ways. There are some possible solutions that can help you fix the Twitch black screen issue and enjoy the stream without any interruption, and you can find them below!

Twitch black screen: Is it a common error?

Twitch is a popular website where content like music, esports, video games, and other things can be live streamed. However, some users may experience a black screen problem when trying to access the website or watch a stream. They may not be able to enjoy their preferred content, which can be frustrating. Thankfully, there are some potential fixes for the Twitch black screen issue.

Check Twitch status

The first thing you should try is to check the Twitch server status, whether it is down or not. If the Twitch black screen issue is caused by the company itself, then there is not much you can do about it but wait for Twitch to figure it out. Here is how to check Twitch status:

1. Open your browser.
2. Go to the official Twitch status page.
3. Scroll down and check the server status.
4. If you see red banners, then there is a server issue, and you can't do much about it.

Twitch

Disable your extensions

Another solution you can try is disabling your browser extensions. Sometimes these extensions cause conflicts with some of the websites, which creates issues like the Twitch black screen problem. Here is how to disable your browser extensions:

1. Go to the "Extensions" tab of your browser. (Usually placed in the settings menu)
2. Remove the extensions that you think are unnecessary.
3. Try refreshing Twitch, and if the issue continues, try removing other extensions too.

Clear your browser cache

If you are still facing the Twitch black screen issue, it is time to try something else and find the fix to your problem. In our third recommendation, we will go with something very common but also pretty useful. You might fix many issues on the internet by clearing your browser's cache.

Google Chrome

1. Open Chrome.
2. At the top right, click More.
3. Click More tools and then Clear browsing data.
4. At the top, choose a time range. To delete everything, select All time.
5. Next to "Cookies and other site data" and "Cached images and files," check the boxes.
6. Click Clear data.

Edge

1. Open Microsoft Edge.
2. In the upper-right corner of the browser, select Settings.
3. Select Settings, then click on Privacy & services.
4. Under Clear Browsing Data, select Choose what to clear.
5. Check Cached images and files and Cookies and other site data, then select Clear now.

Safari

1. Click the Safari tab at the top left of your screen
2. Choose Preferences from the dropdown menu.
3. Click Advanced.
4. Select the “Show Develop menu in menu bar” box.
5. Click the Develop tab from the Safari menu at the top of the page.
6. Click Empty Caches from the dropdown menu.

Restart your router

If none of the steps above worked, maybe restarting your router will help you out with the matter. The smoothest experience on Twitch requires a strong internet connection, but if your connection is poor, the website may occasionally have trouble loading broadcasts or whole pages. Aside from Twitch, see if you can load any other pages first. If neither of them is working, you'll need to unplug and briefly leave your own router unplugged in order to restart it. After that, reconnect the router and attempt to reopen Twitch once your connection has been restored.

Twitch black screen: How to fix it

Apple reportedly prepping a pair of high-end Mac desktops ahead of WWDC

Along with the teaser, Apple has shared a new WWDC23 Power Up playlist that's available to the subscribers of Apple Music. The official playlist contains 25 songs from the likes of Dua Lipa, Post Malone, Miley Cyrus, Coi Leray, The Jonas Brothers, Ed Sheeran, and more. The total length of the curated playlist is 1 hour and 15 minutes.

WWDC 2023 is scheduled to start on June 5 (Monday) at 10 am Pacific Time. While Apple is expected to talk about the incremental updates to its respective software, there have been rumors that it might unveil the long-awaited mixed reality headset as well. With that said, the cryptic message "A new era begins" seems to add some spice to the mix.

For starters, Apple's headset is expected to support millions of existing apps and might get dedicated versions of Final Cut Pro and Logic Pro. According to a report by analyst Ross Young, the headset might come with an insane level of brightness. However, a recent report suggests Apple has made small internal estimates regarding the sales figures of the unreleased headset.

Other than the headset, Apple is expected to announce a 15-inch MacBook Air during the conference. The company could also unveil multiple desktop Macs featuring the M2 Max and yet-to-be-announced M2 Ultra chips. During the conference, Apple will also announce the winners for the 2023 edition of the Apple Design Awards.

via 9to5Mac

Apple shares new WWDC23 Power Up music playlist ahead of its developers conference

Over the weekend, AI-powered makeovers of famous music album covers went viral on Twitter thanks to Adobe Photoshop's Generative Fill, an image synthesis tool that debuted in a beta version of the image editor last week. Using Generative Fill, people have been expanding the size of famous works of art, revealing larger imaginary artworks beyond the borders of the original images.

This image-expanding feat, often called "outpainting" in AI circles (and introduced with OpenAI's DALL-E 2 last year), is possible due to an image synthesis model called Adobe Firefly, which has been trained on millions of works of art from Adobe's stock photo catalog. When given an existing image to work with, Firefly uses what it knows about other artworks to synthesize plausible continuations of the original artwork. And when guided with text prompts that describe a specific scenario, the synthesized results can go in wild places.

For example, an expansion of Michael Jackson's famous Thriller album rendered the rest of Jackson's body lying on a piano. That seems reasonable, based on the context. But depending on user guidance, Generative Fill can also create more fantastic interpretations: An expansion of Katy Perry's Teenage Dream cover art (likely guided by a text suggestion from the user) revealed Perry lying on a gigantic fluffy pink cat.

If you'd like to make your own expanded artwork, you'll need to subscribe to Creative Cloud and download the Adobe Photoshop beta within the Creative Cloud app. Next, open an image and expand the canvas size. Select the entire original image, then inverse the selection so the outer portion of the canvas is selected. If desired, type in a description, then click "Generative Fill." Photoshop will create imagery to fill the space around the original image, and you can flip between various possibilities.

Album covers aren't the only images that have been expanded recently to audience-amusing effect on social media. Aside from obvious expansions of famous paintings like the Mona Lisa, a tweet that expanded several famous meme images garnered almost 10,000 likes on Tuesday.

You can see more expanded album artwork in threads by Dobrokotov and Lorenzo Green on Twitter.

For various understandable reasons, including technical deficiencies in the results, sanctity of art objections, and ethical issues related to utilizing AI-generated artwork, not everyone has been amused or impressed by the results. Browsing quote tweets, it's fair to say that response to the image expansion trend has been mixed between delighted and disgusted. Regardless, Generative Fill has been a visual toy for social media users that keeps on giving.

Source

With outcome supervision, OpenAI trains reward models to provide feedback on the final result the AI gives. With process supervision, the reward model provides feedback at every step of the way, creating a human-like chain of thought.

In its research paper, OpenAI tested both models on a math dataset and found that the process supervision method led to “significantly better performance”. It’s important to note that the process supervision method has only been tested in the area of mathematics so far and that it will take more work to see how it performs more generally.

Explaining the possible outcomes of the process supervision method, OpenAI said:

“If these results generalize, we may find that process supervision gives us the best of both worlds – a method that is both more performant and more aligned than outcome supervision.”

 

It’s still too early to say how much this step-by-step verification will help to address hallucinations more generally, but hopefully, it will because hallucinations are probably the number one issue with LLMs right now. Just this week, a lawyer that had used ChatGPT for his work and submitted false information detailing fake cases that the AI had dreamt up.

OpenAI has not given a timeline for how long it will take to implement process supervision in ChatGPT which is available to the public. It’s still in the research phase and needs to be tested on general information.

While initial results are good, OpenAI does mention that safer methods can incur reduced performance called an alignment tax. The results show so far that process supervision doesn’t incur this tax while working on math problems but we don’t know what will happen on more general information.

Source

Earlier this year, Twitter decided to shut down third-party apps or impose severe restrictions on their usage. This move frustrated many developers, as they relied on alternative apps for enhanced user experiences. It appears that Reddit may be following a similar path, raising concerns within the developer community.

Apollo has gained popularity as one of the most Reddit clients. However, the future of third-party apps now hangs in the balance. Christian Selig shared his interactions with Reddit regarding the updated API pricing, which revealed a significant cost increase.

According to Selig, Reddit informed him that the new API cost would be $12,000 for 50 million requests. Given that Apollo generates around 7 billion requests per month, the estimated monthly expense for API access would reach a staggering $1.7 million, equivalent to $20 million annually.

Selig further highlighted the drastic difference in costs between Reddit and other platforms. For instance, he mentioned that 50 million API calls with Imgur cost him $166. In comparison, Reddit's proposed pricing is approximately 20 times higher.

The API costs pose a significant challenge for third-party developers. Even if he were to rely solely on subscription users, the average Apollo user's monthly usage would cost $2.50, more than double the current subscription fee.

Despite the situation, Selig is determined to find a solution and continue supporting Apollo. However, Reddit has indicated that it could be more flexible regarding the API pricing, leaving Selig and other developers in a precarious position.

As the situation unfolds, the developers and Reddit must find a mutually beneficial solution that ensures the continued availability and success of these popular third-party Reddit clients.

Source: Reddit Thread

Reddit wants to make a 'million dollars' from API following Twitter

In the world of processor making, there are two types of companies. Fabless and foundry.

Fabless companies only design chips like CPU, GPU and others. While foundries make them.

In terms of foundries with latest cutting edge tech, there are only 3 big names in the whole world. Intel, Samsung and Taiwan based TSMC.

In terms of fabless companies, there are many of them. They include, AMD, Nvidia, Qualcomm, MediaTek and a lot more.

But which of the three makes chips for these fabless companies. Currently, most of the top chips by fabless companies are made by TSMC, which is a market leader. Second in the list is Samsung, which hardly gets any customers these days due to it being behind TSMC it terms of abilities.

But what about Intel. For years, Intel has made chips for only itself and not others. That changed some time ago when Intel announced that it wants to enter the foundry race too. That means, Intel wants to create CPUs and GPUs for not only for Intel’s own consumption but for other fabless companies as-well.

In-fact, today itself, Intel’s General manager of Intel Foundry Services and VP Stuart Pann has shared a detailed post about the company’s vision moving ahead in creating chips for other companies.

Looks like Intel has got even Nvidia interested too.

Nvidia Talks Intel Made GPU

Currently, world’s largest computer trade show Computex is going on in Taiwan. Companies are announcing and showcasing their new computer related products with a great amount of enthusiasm. A big amount of tech journalists and enthusiasts are there too.

In it, Nvidia’s CEO Jensen Huang was having a small show where a lot of top journalists were involved. Tom’s Hardware reports that during that conference, Nvidia’s Jensen revealed something really interesting.

Jensen said that Nvidia has used Samsung made GPUs in past (Nvidia RTX 30 series) and is looking at a possibility in using Intel made GPUs in future too.

In-fact, Jensen said that Nvidia has received a test GPU from Intel, using its next-gen process. Quoting him:

“You know that we also manufacture with Samsung, and we’re open to manufacturing with Intel. Pat [Gelsinger] has said in the past that we’re evaluating the process, and we recently received the test chip results of their next-generation process, and the results look good,” Huang said.

Reasons Why Nvidia Is Considering Intel

As interesting it is. One has to understand why Nvidia is considering Intel in the first place. Especially because Intel is now Nvidia’s rival in graphics card making.

From decades, Nvidia has used GPUs made by TSMC. With the company rarely using other foundries. So there has to be some big reason for Nvidia to consider Intel. There are not one, but two reasons behind it. Price and tech.

In prices, TSMC is consistently increasing the prices of the silicon wafers made by them. Those unaware, computer chips are made from silicon wafers, which are later cut into proper CPU and GPU shapes after they are made, to put it very simply.

TSMC is consistently increasing the prices of every new process, that too in big numbers.

For example, a single 7nm wafer by TSMC used to cost $10,000 in 2018. Move to 5nm two years later, the price has been raised to $16,000 per wafer.

For 3nm, TSMC is demanding a massive $20,000 a wafer. That 3nm process isn’t a completed one, either. TSMC is working on an enhanced version of 3nm, called N3E, which can cost even more. DIGITIMES reports (translated) that TSMC has no intention of cutting prices even during lack of demand.

So Nvidia wants to look elsewhere. Anywhere where they can order chips at a cheaper cost.

The second reason Nvidia could be considering Intel is that the company’s next-gen process is looking great.

WikiChip Fuse reports that in terms of transistor density, Intel 4 process is comparable to TSMC’s 3nm process. Which is an outstanding improvement by Intel if it turns out to be true.

It means Intel is moving fast with innovation and is beating its rivals quickly. Something which Intel was lagging massively from a few years.

Conclusion

What’s interesting is that originally, Intel wanted to sign up other companies for its 2nm process and more. But it seems to be in a hurry and wants to start with 3nm itself.

Either way, a possibility of Nvidia using Intel is something worthy of a notice. But will Nvidia actually use Intel made GPU, we cannot say for sure. Who knows, Jensen could be talking about Intel only to bargain for better prices with TSMC.

Another possibility is that geopolitical issues means that Nvidia has to make sure that in-case an invasion of Taiwan does happen, it has other options available with them.

We do believe there should be more foundry companies out there to provide a good competition to TSMC, who is almost a monopoly now. So Intel, a big name in itself, fits the idea perfectly here.

Nvidia Could Use GPU Made By Intel In Its Graphics Cards

The deadly non-integrated I/O shield (Image credit: Windows Central)

Have you built yourself a shiny new gaming PC in recent times? Did you shed actual blood while trying to install your motherboard? If the answer is no, you probably got lucky or bought one of the best motherboards with an integrated I/O shield. I wish I had.

I finished recently reassembling my desktop gaming PC to put my new ASRock-flavored Intel Arc A770 to work and nearly severed the top of my finger clean off. Trying to install the I/O shield into the case, my finger slipped ever-so-slightly, and then I felt it. Like a hot knife through butter, the sharp edge of the I/O shield sunk cleanly into my finger.

I've never seen so much blood come from my finger in my entire life.

Cheap metal, dangerous to use

If the motherboard you want to buy looks like this, walk away. Your fingers will thank you.  (Image credit: Windows Central)

I get it. Businesses will always cut every corner imaginable to maximize their profits. But it also shouldn't come at the expense of the customer's well-being. And this isn't a new phenomenon. It doesn't take much time searching places like Reddit to find scores of injured PC builders.

I won't post pictures of my bloodied digit, but it wasn't good. It hurt. Bad. Overall, it took about two days to finally stop bleeding because I kept re-opening it every time I changed the dressing, and fortunately, it hasn't left a scar or anything.

But why is this even a thing? When you compare the quality of an integrated I/O shield with the garbage we get when it's not, why is this still happening in 2023? The metal used in these things is of poorer quality than cans of processed food. It's flimsy, it's nasty, and it's sharp as hell. Why?!

It also makes choosing a motherboard for your PC even more important. Beyond the specs, design, and expansion slots, you need to ensure you're getting one with an integrated I/O shield. Please don't make my mistake.

An easy fix

Every motherboard needs to be shipped like this.  (Image credit: Future)

There's an easy fix here; just stop shipping these things. There is absolutely no benefit to anyone in not having an integrated I/O shield. All it causes is pain and misery.

Motherboards are hardly cheap items, but if it meant adding $5 to the price of each of them to do away with these weapons of mass finger destruction, I don't think a single PC builder in the world would mind.

Integrated I/O shields look far superior, for one, but they also don't take big lumps out of your flesh. It's time to send these sheets of razor-sharp metal to the dumpster once and for all.

A plea to PC motherboard makers: Stop making us BLEED!

The original Chromecast hits end of life after a decade of service

Below are the full specifications of this thing. I have bolded my configuration where relevant.

The Intel NUC 13 is offered in a few different configurations on the GeekNUC store, with our i7-1360P Raptor Lake chip and either a 512GB SSD and 16GB of RAM or 1TB SSD and 32 GB of RAM option, or with the i5-1340P chip with only 512GB SSD and 16GB of RAM to choose from.

You can also get it without memory, storage, or an operating system license for $99 less, which affords the option of putting in better components such as different RAM sticks, a better NVMe drive, and your own choice of operating system. Even if you decide to go with your own RAM chips, unfortunately you're stuck with DDR4 with this system, DDR5 support starts with the NUC 13 Pro Extreme series, which also costs double.

If you're not familiar with Intel NUC Mini PCs, they come in a "Tall" or "Desk" variant, the main difference is that the "Desk" variant does not allow for an additional 2.5-inch SSD to be added because it misses the 2.5-inch SSD compartment.

The packaging definitely has that familiar Intel design look about it due to the colour. Once you pull the top off you are greeted with the Mini PC sitting in a laminated cardboard compartment covered with thin plastic that can be peeled off.

Upon removing the Mini PC, safety documentation, and an Intel Core i7 sticker and then removing the cardboard compartment "shelf", there are 3 cardboard compartments containing the VESA mounting plate, power supply and other items that are listed below.

What’s In The Box

• 3 x User Manual
• 1 x VESA Mounting Plate
• 1 x Intel Nuc 13 i7 Mini PC
• 1 x Power Supply
• 1 x AC Cable
• 6 x Screws

Omissions

Intel did not feel it necessary to include a v2.1 HDMI cable, so this is something you'll have to purchase yourself, ensuring of course that it supports the modern standard and you don't get one that is only 2.0b compatible.

Design

The Intel NUC 13 design is a familiar one for Mini PCs and actually 255g heavier than the GEEKOM Mini IT12 we looked at recently, it definitely has a good heft to it and I was initially surprised at the overall packaged weight. Coming in at just under a kilo, it's not heavy, but it certainly isn't light either.

The case is entirely black and plastic, and the top cover has a light textured feel to it, so it isn't a glossy fingerprint magnet. It lacks the premium finish I saw on the GEEKOM Mini IT11 and IT12. All four corners are rounded, and it has a balanced feel to it, so it is comfortable to hold and move around.

Intel used up to 95% recycled materials with the packaging and the NUC 13 Pro case itself is made up of recycled plastics, so even though it has a plastic "look" to it, we really should be praising the efforts of Intel to be environmentally conscience with the manufacturing of it.

The front of the NUC 13 Pro includes two USB 3.2 Gen 2 type A ports and a 3.5mm headphone jack port. Unfortunately, there's no Type C port on the front, which would have been nice, but you can find the two Thunderbolt 4, which use the Type C connector port, around the back.

As far as looks go, and as previously mentioned, it's completely black, and it's not a fingerprint magnet thanks to the slightly textured plastic surface on the lid which you can see in the above photo, it also has the "intel nuc" logo stamped in the bottom left corner of the lid. Essentially, it's a plastic shell covering a metal frame. It doesn't feel cheap, and isn't flimsy either, and even when the base is removed it remains sturdy.

Accessing the Intel NUC 13 Pro is as easy as unscrewing the bottom plate with the four Philips head screws located in the center of each rubber foot, a small hobby screwdriver is enough for the job, but you may want to use a flat-head screwdriver to prise the plate free, taking care not to damage the ribbon which connects to an SSD port that allows for an optional full sized 2.5-inch SSD to be screwed to the backplate. Helpfully, the backplate screws do not come loose from the rubber feet, so you won't lose them either.

As you can see from the above images which can be enlarged when clicked on, there is ample room to manage the SODIMMs and Kingston 512 GB PCIe 4.0 NVMe SSD, so you can swap it out for something else if you want. The Intel NUC 13 Pro's motherboard includes an option to add a second NVMe SSD, as long as it is only 42mm long.

The backplate only fits back in one way, this is to ensure that the NVMe is always connected to the heat spreader. There's also an option to secure a 2.5-inch SSD with four screws if the four rubbers surrounding the device information sticker are removed, this would be required since there is no "sled" to hold such an SSD in place.

Usage

On first boot, you are prompted to complete the setup of Windows 11 Pro, meaning you do not have to fork out for a license, which is nice. After the setup is completed, I am happy to report that the NUC 13 Pro is not loaded with any third-party bloatware. It ships with Windows 11 22H2, so updates were relatively minimal.

The Intel NUC 13 Pro also supports DisplayPort 1.4a through the Thunderbolt 4 port so I can connect my Samsung Odyssey Neo G9 using a USB 4 to DisplayPort cable and utilize the maximum native resolution for my screen, which is 5120x1440 @ 240Hz.

It's physically possible to directly attach four screens to the Intel NUC 13 Pro using the two HDMI and the two Thunderbolt 4 ports if you wanted. Regarding connectivity, there is a USB 2.0, and a 3.2 gen 2 port along with a barrel port for power on the back. Around the front, there are two more USB 3.2 Gen 2 ports and a port for a 3.5mm headphone jack.

For audio output, I linked my Edifier 360DB over Bluetooth, and I did not experience any significant audio delays. It took a few reboots for the Intel NUC 13 Pro to recognize and connect my speakers automatically, the first few times I had to disable and re-enable Bluetooth for the speakers to get connected.

As you can see from the two images above, you can affix a Kensington lock on one side, and both sides contain a mesh grill for air cooling. You may have noticed there is no card reader present, although USB card readers are relatively inexpensive, it would have been nice to have™.

Benchmarks

Before I started running benchmarks, I ensured that Windows 11 and drivers were up to date. At the time of testing, the NUC 13 Pro was running Windows 11 Professional 22H2 build 22621.1778 and I also upgraded to the latest Intel driver which is v31.0.101.4369 (May 2023).

During one of the many restarts due to updates, I was also prompted to install a BIOS update that was released on May 23 that had the following changelog:

This download record contains options for updating the Intel® Aptio* V UEFI Firmware BIOS for Intel® NUC 13 Pro - NUC13AN

With that out of the way, and because people like that sort of thing, I ran some benchmarks and compared it to my main PC which I built in 2019 and benchmarked in December 2022.

My Main PC consists of the following:

• Intel Core i9-9900K (stock)
• Gigabyte Z390 AORUS PRO WIFI (BIOS revision F12)
• 64GB DDR4-3200 G.Skill Ripjaws (16-18-18-38) dual channel
• Samsung 980 1TB NVMe
• KFA2 RTX 2070 SUPER (Nvidia driver 527.37)
• Windows 10 22H2 build 19045.2311

For the benchmarks, I used 3DMark, PCMark 10, Geekbench, and Cinebench. 3DMark Time Spy tests gaming capability with DX12 graphics performance.

PCMark tests are a mix of CPU and real-world productivity tests, such as using an office suite, web browsing, light photo/video editing, and making conference calls. Cinebench stresses the entire CPU as it is a multi-threaded rendering test. Finally, Geekbench is a synthetic benchmark that is great for a quick look at the potential performance across a wide range of workloads.

As you can see, the i7-1360P processor is starting to catch up with the four-year older i9-9900K in terms of compute score, the only thing that lets it down is the weak graphics prowess of the Iris Xe GPU.

I also tested the SSD's capability using AS SSD and CrystalDiskMark.

Despite running all of the above benchmark tests, the NUC 13 Pro did not get hot to the touch, but it did get a little warm, and there were no annoying noises coming from the single fan that cools the unit.

For kicks and to blow off a bit of steam, I decided to give Quake Champions a go to see if it would be playable, and after toning down the screen resolution to 3840 x 1080 with everything set to "Low" (which was auto-detected anyway) I can say that it was playable, just about. And here we are talking about a game that first went into closed beta in 2017 using a hybrid game engine made up of id tech and Saber tech.

Conclusion

One thing that is really clear about the Intel NUC 13 Pro is that it isn't a gaming PC, you will not be able to enjoy graphically intensive games on it, but it absolutely is suited as an office workstation, or perhaps a good solution for a student or office worker without a permanent desk affording the ability to pack this away after every use. This thing isn't taking up much room, you can even screw it to the back of a screen if it has VESA support with the included mount plate.

When it comes to Mini PCs, the market is saturated with options like the BEELINK SEI12 with its mobile-class 12th gen Intel i5-1235U being one such example, but that option does not include Thunderbolt 4, dual channel memory, or even a Kensington lock and more. It's the lowest denominator of a Mini PC and when you're spending hundreds of dollars to replace the job of a full-sized PC, you're going to want it to replicate it as much as possible aside from the obvious lack of GPU prowess.

The NUC 13 Pro also has the cheaper i5 entry point, and a beefier i7 model in two flavors with double the memory and storage costing $100 more, and there's even a barebones option where you can put in your own choice of memory, NVMe SSD and OS.

As I said earlier, the decision ultimately comes down to what you're willing to pay for the options you need. This Mini PC will let you connect up to four displays, whereas cheaper solutions will be limited to two screens.

Staring at $649 for the barebones i5-1340P variant, the NUC 13 Pro is not cheap by any means, but it contains the newest of the newest 13th gen Intel processor chipset with PCIe 4.0 connectivity. Aside from not being able to seriously game on it, you are still getting a powerful machine inside a tiny body.

But wait, there's more

For a while at least, there's a coupon getting you $100 off at GeekNUC when you enter the coupon code NUC13PRO100 which you could apply to the beefier variant, but happily, it applies to all possible variants on their store.

If you prefer to shop through the benefits of Amazon for free delivery and returns, you will be able to save a bit of money on our variant; our contact at GeekNUC let me know they are running a promotion through their official Amazon page starting on May 30th that gets you an extra 10% off the price when you enter the 9FUHM9TC coupon upon checkout, bringing it down to $809.10 or $899.10 for the 32 GB/1 TB option or $719.10 for the barebones variant. This promotion ends on July 30th.

Our stories may contain affiliate links for products/apps where Neowin is paid an affiliate fee if you complete a purchase via those links.

Intel NUC 13 Pro review: Packing 13th-gen Raptor Lake power inside a tiny body

A bacterial protein found in English oak buds could be used to more efficiently process the rare earth elements needed for different technologies.

Rare earth elements are used in products such as smartphone and TV screens, electric vehicle motors and wind turbines. Natural deposits of these elements can be found and mined in multiple countries. But the industrial separation facilities necessary for processing them use an energy-intensive method that typically requires hundreds of steps and uses many toxic chemicals – something that could be simplified through the natural shortcuts provided by a certain class of bacteria.

Joseph Cotruvo Jr. at Pennsylvania State University and his colleagues discovered that the bacterium Hansschlegelia quercus contains a protein that can help differentiate between lighter and heavier types of rare earth elements. They isolated the bacterium from English oak buds while examining hundreds of such bacterial proteins – building on their previous research showing how proteins found in the metrolyph class of bacteria can help separate rare earth elements from non-rare earth elements.

The team found that if the bacterial protein binds to a lighter rare earth element, it is more likely to also bind to another unit of itself. If the protein binds to a heavier rare earth element, it is more likely to stay single.

This discovery allowed the researchers to show how the protein can help separate out neodymium and dysprosium – light …

and heavy rare earth elements respectively that are used in permanent magnets – in a single chemical step at room temperature without relying on toxic chemicals.

The approach could help process rare earth elements that have either been mined from the ground or are being recovered from recycled tech products.

The work is a “proof-of-concept” approach that still faces significant challenges in being able to process rare earth elements on a large industrial scale, says Jason Love at the University of Edinburgh in the UK. “A significant issue with the biochemical approach is to ‘manufacture’ enough protein to be able to carry out these separations at scale.”

Journal reference   Nature DOI: 10.1038/s41586-023-05945-5

Science

These innovative plugins, designed by third-party developers, bring enhanced functionality to ChatGPT by granting it access to the internet and specific services.

What are ChatGPT Plugins?

ChatGPT Plugins serve as valuable tools to augment the functionalities of ChatGPT. Developed with a core emphasis on safety, these plugins allow ChatGPT to access real-time information, perform computations, and utilize third-party services.

With the incorporation of Plugins, ChatGPT transcends its initial role as a simple chatbot and transforms into a dynamic platform for interaction and exploration.

By integrating Plugins, OpenAI aims to position ChatGPT as a strong competitor to other AI chatbots like Google Bard and Microsoft Bing AI.

ChatGPT Plugins are third-party additions to popular chatbot by OpenAI - Image courtesy of OpenAI

How to use ChatGPT Plugins?

To make the most of ChatGPT Plugins, follow these simple steps:

1. Get ChatGPT Plus: You need a ChatGPT Plus subscription to access Plugins
2. Open ChatGPT settings: Go to the bottom left corner and click on "Settings"
3. Enable beta features: Within the settings, select "Beta Features" to access experimental options
4. Toggle the Plugins setting: Enable the Plugins functionality within ChatGPT
5. Start a new chat: Enter the chat window to interact with ChatGPT
6. Switch to GPT-4 model: Select the GPT-4 model to optimize Plugin performance
7. Access the Plugin store: Open the dropdown menu and choose "Plugins" to enter the Plugin Store
8. Choose and install Plugins: Browse the available Plugins, select the ones you want, and install them

By following these steps, you can enhance your ChatGPT experience with Plugins, gaining access to up-to-date information and specialized services.

A glimpse into the expanding plugin ecosystem

As of May 2023, there are approximately 85 ChatGPT Plugins available on the dedicated Plugin store. This number is expected to grow rapidly in the coming months as developers become more familiar with OpenAI's guidelines and the platform itself.

The development of these Plugins aims to seamlessly integrate existing websites and services into the ChatGPT platform, further expanding its capabilities.

What are the best ChatGPT Plugins out there?

With such a diverse and expanding ecosystem, ChatGPT is poised to offer a comprehensive and user-friendly experience.

Our recommendations on the best ChatGPT Plugins are as follows:

• Wolfram
• ChatWithPDF
• Instacart
• Video Insights
• Kayak
• Link Reader
• KeyMate AI Search
• OpenTable
• Golden

Wolfram - Empowering ChatGPT with computation and knowledge

The Wolfram Plugin is a standout among the array of ChatGPT Plugins. It provides access to computation, math, curated knowledge, and real-time data.

By leveraging the power of WolframAlpha and Wolfram Language, this Plugin offers users accurate and comprehensive answers.

Whether users are seeking information on complex topics or the latest data trends, the Wolfram Plugin proves to be an invaluable resource.

Wolfram serves as one of the best ChatGPT Plugins out there which enhances chatbot's capabilities - Image: Wolfram Alpha

ChatWithPDF - Seamlessly interacting with PDF files

ChatWithPDF is another remarkable Plugin that allows users to interact with PDF files effortlessly. By simply pasting the URL link of any PDF file, users can ask questions about its content within ChatGPT.

This Plugin proves particularly useful when dealing with large and complex documents, such as textbooks, reports, contracts, books, and papers. With ChatWithPDF, extracting vital information from PDF files becomes a seamless process.

Instacart - Simplifying grocery shopping and meal planning

For those seeking assistance with grocery shopping and meal planning, the Instacart Plugin is an excellent choice. This Plugin enables users to ask questions about recipes, ingredients, products, and prices.

With access to over 1.5 million products across more than 1,000 stores, the Instacart Plugin streamlines the shopping experience.

Additionally, once users receive personalized meal suggestions, the Plugin conveniently provides a clickable link to a pre-filled shopping basket.

Instacart saves you from the hassle of meal planning and shopping - Image: Instacart

Video Insights - Unveiling the essence of YouTube videos

The Video Insights Plugin brings a new dimension to ChatGPT's capabilities. By pasting the URL of any YouTube video into the prompt, users can request a short summary of the video or ask questions about its content. Whether users seek inspiration for new YouTube video ideas or need a concise overview of a video's content, the Video Insights Plugin is a valuable tool. It should be noted that the Plugin currently supports videos up to 10 minutes in length.

Kayak - Optimizing travel planning

For travel enthusiasts, the Kayak Plugin offers a convenient solution. This Plugin allows users to inquire about flights, hotels, car rentals, and activities, providing personalized recommendations based on their preferences and Kayak's extensive database.

By utilizing the Kayak Plugin, users can quickly and conveniently find the best deals and book their dream trips.

Link Reader - Enriching prompts with Context

The Link Reader Plugin serves as a valuable tool for enriching prompts with additional context. By pasting any link into ChatGPT, users can utilize this Plugin to read and extract the content from the provided link.

This proves especially helpful when the ChatGPT web browsing feature encounters limitations or restrictions. The Link Reader Plugin enables seamless connectivity between ChatGPT and the internet.

KeyMate AI Search - Streamlining web browsing

Another plugin that enhances web browsing capabilities is KeyMate AI Search. By utilizing Google's search API, this Plugin scans and summarizes the top search results for a given keyword.

It analyzes text summaries from the first search page to provide ChatGPT with up-to-date responses.

KeyMate AI Search offers a valuable alternative to ChatGPT's built-in web browsing feature, particularly when some websites restrict access to non-human web visitors.

OpenTable - Discovering and booking restaurants

For food enthusiasts looking to explore new dining experiences, the OpenTable Plugin comes to the rescue. With this Plugin, users can ask ChatGPT to show them restaurants based on their location, cuisine preferences, budget, or other specific criteria.

Drawing from OpenTable's extensive database of over 60,000 restaurants worldwide, this Plugin provides users with tailored recommendations and convenient links for booking reservations.

OpenTable allows you to get outstanding dining recommendations using ChatGPT - Screenshot: OpenTable

Golden - Accessing financial and business data

The Golden Plugin opens up a world of financial and business data to ChatGPT users. With this Plugin, users can easily retrieve company statistics and other market information. Golden Plugin acts as a gateway to more recent sources of financial data, empowering users with valuable insights.

With their ability to enhance ChatGPT's functionality by connecting it to the internet and specific services, these Plugins offer users expanded opportunities for interaction, knowledge acquisition, and convenience. As the Plugin ecosystem continues to grow and improve, ChatGPT is poised to become an even more indispensable tool for various tasks and scenarios. So why not explore the world of ChatGPT Plugins and unlock new possibilities with ease?

Source

Speaking before a Senate Banking Committee hearing, Paul Rosen, who oversees investment security at Treasury, said officials are working on curbing investment from the U.S. that "comes with know-how and expertise to certain specific sectors and subsectors such as advanced semiconductors, artificial intelligence and quantum computing," citing China and its military in particular.

Reuters reported in February that the Biden administration plans to ban investments in some Chinese technology companies and increase scrutiny of others, three sources said, as part of its plan to crack down on the billions that American firms have poured into sensitive Chinese sectors.

China hawks in Washington blame U.S. investors for transferring capital and valuable know-how to Chinese tech companies that could help advance Beijing's military.

Separately, Republican Senator Bill Hagerty asked about efforts to restrict the supply of U.S. origin goods to Chinese telecommunications company Huawei.

Exports to Huawei currently require a license. Asked about revoking these licenses, Commerce Department assistant secretary Thea Rozman Kendler said, "We do not have a draft rule at this time," but added that "We are under deep analysis of this issue."

Officials are closely scrutinizing a wide range of exports to China and said that last year they denied or took no action on a quarter of requests in order to stop sales that would advance Beijing's militarization.

In 2022, 5,064 export and re-export license applications were reviewed and about 26% were denied or returned without action, Kendler said in written testimony ahead of the Senate Banking Committee hearing.

Reporting by David Shepardson and Daphne Psaledakis in Washington, and Karen Freifeld in New York; Writing by Chris Sanders; Editing by Daniel Wallis

Source

Many of these new technologies chase the dream of optimized slumber. They promise to help tailor our sleep schedules to fit around our social lives, help us sleep for longer or even skip a night's sleep altogether.

Here's how technology is permeating our sleep, and what the future holds.

Time to wake up

Sleeping pills have recently been joined by a wave of wakefulness drugs, purportedly safer and more powerful alternatives to caffeine. It seems that they work best on people who are already sleep deprived and don't have a huge effect on those who are already well rested.

Modafinil is touted for its cognition enhancing effects (especially in sleep-deprived people) and can supposedly keep people awake and alert for several days at a time. Some scientific studies are showing that this may indeed be the case, although results are mixed, with other research showing the effects are similar to caffeine.

The drug was developed to help people with narcolepsy but some have started using it for its focus-enhancing effects. It is a controlled drug (prescription only) in most countries. People who use it for cognitive enhancement or wakefulness are buying it on the black market or getting it from friends who have a prescription.

Modafinil is popular with students—in 2020, Loughborough University researchers found that, of 506 students surveyed at 54 UK universities, 19% had taken cognitive enhancement substances.

But people who take them for non-medical purposes are risking their health. Studies of the safety of these drugs do not consider this type of use. We don't know what using these drugs to stay awake for long periods of time does to people's bodies. But we do know that disrupting your sleep pattern (for example, shift work) is linked to health problems such as diabetes and cardiovascular disease.

Recent studies suggest some people are combining sleep and wakefulness pills to manage their body rhythms and optimize their sleep or unwind after a day of hard work. The effects of taking wakefulness pills with other drugs is largely unknown.

In the UK sale or supply of a prescription-only or unlicensed medicine is a criminal offense. Whereas in the US, even possession of stimulants without a prescription is a crime.

Smart sleep

Many people already use smart watches, smart jewelry and fitness bands to track their sleep—for example, alarms that wake people up at the optimal point in their sleep cycle and motion sensor apps that analyze sleep patterns.

New ways of tracking sleep could soon include donning a pair of pajamas embedded with sensors to track changes in posture, respiratory and heart rate, or hugging a robot pillow, whose algorithm creates a breathing pattern to mimic and help you fall asleep.

Meanwhile, care robots have already been trialed in Japan to test whether they could help older people sleep better. Designed to watch over residents at night in care homes, they give staff information on how well the residents are sleeping and let them know if anyone goes for a nocturnal wander.

In your dreams

Dream management technologies are in much earlier stages of development. Scientists believe that sensory stimulation technologies and devices, such as virtual reality visors, could be used for sleep engineering. This new science involves exposing the sleeper to sensory stimuli, such as clicking sounds and vibrations, at specific times in the sleep cycle. The aim would be to improve sleep quality, enhance memory and even treat post-traumatic stress disorder (PTSD).

As for the prospects of "reading" our dreams, progress is being made on this front too. Scientists have taken the first steps towards dream interpretation by measuring brain activity during sleep and using AI to decode visual imagery. Participants in a 2013 study were asked to report the imagery from the dreams after sleeping inside an MRI scanner. Researchers compared scans from people viewing the same types of images while awake and the results showed matching patterns of brain activity.

Nightmare technology

But there is a dystopian side to this story. The technology we already have—electric light, smart phones, streaming services—can be disastrous for our sleep.

For example, a recent study in the US found that college students often sleep with their mobile phone in bed with them, which means a call, software update or app notification can disturb them. Watching TV or playing video games in bed and staring at our tablets and mobile phone screens into the night has become the norm for many. It can lead to poor sleep and knock our sleep cycles off kilter.

Growing numbers of people are seeking treatments for new sleep conditions such as orthosomnia—the obsessive quest for perfect sleep, similar to an unhealthy preoccupation with nutrition. Some people become so concerned about improving their sleep metrics that it is actually giving them insomnia.

There's still so much we don't know about sleep, and new technology is changing our sleep faster than scientists can keep up with. One thing seems almost for sure: sleep and technology in western society are becoming entangled like never before.

Provided by The Conversation

Source

In a blog post, Microsoft says this latest version of its mixed reality app development toolkit uses the OpenXR standard. That means developers that use this new tool will be able to make apps that will not only work on Microsoft's HoloLens device, but other similar headsets as well.

The Microsoft Mixed Reality Toolkit 3 is also compatible with the popular game development engine Unity, It uses that engine's XR Management System and XR Interaction Toolkit so the tools will be more compatible with new versions of Unity.

The new version of the toolkit will have these specific improvements:

• Faster performance: Developers can improve the performance of HoloLens 2 and resource-constrained mobile platforms.
• Smaller memory footprint: By utilizing OpenXR, developers can better reduce the memory footprint of their cross-platform apps.
• Modern user interface: Support for new Unity features like Unity 3D Canvas reduces the time it takes to code and launch apps into production.

Microsoft also announced today that the Azure Communication Services Calling SDK for Unity will have a public preview launch before the end of the second quarter of 2023. Microsoft stated:

Not only can developers benefit from all the features inherent in the ACS Calling SDK for UAP, but they also can leverage Unity Editor. This gives developers access to a visual interface that significantly reduces the time it takes to create, design, and build HoloLens 2 applications.

In April, Microsoft released the free Windows 11 update for the HoloLens 2. The company has said it is committed to updating the HoloLens 2, but there's no word on if the company plans to release a next-gen version of its own expensive mixed reality headset.

Source

Advances in artificial intelligence (AI) are complicating publishers’ efforts to tackle the growing problem of paper mills — companies that produce fake scientific papers to order. Generative AI tools, including chatbots such as ChatGPT and image-generating software, provide new ways of producing paper-mill content, which could prove particularly difficult to detect. These were among the challenges discussed by research-integrity experts at a summit on 24 May, which focused on the paper-mill problem.

“The capacity of paper mills to generate increasingly plausible raw data is just going to be skyrocketing with AI,” says Jennifer Byrne, a molecular biologist and publication-integrity researcher at New South Wales Health Pathology and the University of Sydney in Australia.

“I have seen fake microscopy images that were just generated by AI,” says Jana Christopher, an image-data-integrity analyst at the publisher Federation of Eeuropean Biological Societies Press in Heidelberg, Germany. But being able to prove beyond suspicion that images are AI-generated remains a challenge, she says.

Language-generating AI tools such as ChatGPT pose a similar problem. “As soon as you have something that can show that something’s generated by ChatGPT, there’ll be some other tool to scramble that,” says Christopher.

A stream of papers

Anna Abalkina, a social scientist at the Free University of Berlin and an independent research-integrity analyst, suspects that there might be a delay in these AI tools becoming more apparent in the academic literature because of the length of the peer-review process. Perhaps in the next few months, “we will see the first stream of papers”, she says.

Byrne, Christopher and Abalkina were participants at the UNITED2ACT summit last week, which was convened by the Committee on Publication Ethics (COPE), a non-profit organization focused on ethics in academic publishing based in Eastleigh, UK, and the International Association of Scientific, Technical and Medical Publishers (STM), based in Oxford. The summit brought together international researchers, including independent research-integrity analysts, as well as representatives from funding bodies and publishers.

“It was the first time we had a group of people come together and co-create a set of actions which we’re going to take forward to combat this problem,” says Deborah Kahn, a trustee of COPE and a research-integrity consultant, based in London. The group intends to publish its joint action plan soon.

When it comes to detecting paper-mill works, “there is absolutely an additional challenge which is posed by synthetic images, synthetic text, et cetera”, says Joris van Rossum, programme director for STM Solutions, a subsidiary of STM. “There is a general realization that there is the potential of screening becoming more difficult,” he says.

AI assistance

Kahn says that, although there will undoubtedly be positive uses of AI to support researchers writing papers, it will still be necessary to distinguish between legitimate papers written with AI and those that have been completely fabricated. “We have to really look at how we identify those things, and how we make sure that people have actually done the research. And there are various ways we can do that,” she says.

One strategy discussed during the summit was to require authors to provide the raw data from experiments, potentially with digital watermarks that would enable publishers to confirm that those data are genuine.

Currently, requirements for submitting raw data vary significantly between publishers, says Christopher. Establishing a uniform set of requirements for the submission of raw data across publishers, taking into account differences between fields of research, could therefore be helpful, she says.

Sabina Alam, director of publishing ethics and integrity at Taylor & Francis, a publisher based in Abingdon, UK, agrees but says that such standards will take time to implement. “I can’t imagine it being an overnight flip, because the reality is many institutions don’t actually have the resources to offer data-management infrastructure,” she says. “We don’t want to penalize actual research.”

Sharing information

The summit also discussed other strategies for tackling the problem of paper mills more broadly, including organizing an awareness day or week for researchers, as well as identifying ways for publishers to share relevant information on suspected paper mills — for example when publishers simultaneously receive submissions — without breaching data-protection rules.

STM is continuing to develop its own paper-mill detection software, while also collating resources on similar tools available elsewhere through its integrity hub. The apparent rise in paper mills increases demand for such techniques — both for detecting fake papers at the point of submission and for identifying those that are already published.

Taylor & Francis is among the publishers that are making use of such tools, and Alam says that a growing number of ethics cases — instances of potential misconduct that are flagged for further investigation — are being escalated to her team. Roughly half of these cases are because of paper mills, according to Alam. Her team saw the number of ethics cases increase more than tenfold from 2019 to 2022 — and so far this year, there have been almost as many cases as during the whole of 2022. “It seems to have been commercialized and scaled up,” she says.

doi: https://doi.org/10.1038/d41586-023-01780-w

Source

A few months ago, while meeting with an A.I. executive in San Francisco, I spotted a strange sticker on his laptop. The sticker depicted a cartoon of a menacing, octopus-like creature with many eyes and a yellow smiley-face attached to one of its tentacles. I asked what it was.

“Oh, that’s the Shoggoth,” he explained. “It’s the most important meme in A.I.”

And with that, our agenda was officially derailed. Forget about chatbots and compute clusters — I needed to know everything about the Shoggoth, what it meant and why people in the A.I. world were talking about it.

The executive explained that the Shoggoth had become a popular reference among workers in artificial intelligence, as a vivid visual metaphor for how a large language model (the type of A.I. system that powers ChatGPT and other chatbots) actually works.

But it was only partly a joke, he said, because it also hinted at the anxieties many researchers and engineers have about the tools they’re building.

Since then, the Shoggoth has gone viral, or as viral as it’s possible to go in the small world of hyper-online A.I. insiders. It’s a popular meme on A.I. Twitter (including a now-deleted tweet by Elon Musk), a recurring metaphor in essays and message board posts about A.I. risk, and a bit of useful shorthand in conversations with A.I. safety experts. One A.I. start-up, NovelAI, said it recently named a cluster of computers “Shoggy” in homage to the meme. Another A.I. company, Scale AI, designed a line of tote bags featuring the Shoggoth.

    RLHF shoggoth @scale_AI tote bags incoming

    courtesy @TimTheSloth pic.twitter.com/RIK4bjHByu

    — Alexandr Wang (@alexandr_wang) March 27, 2023

Shoggoths are fictional creatures, introduced by the science fiction author H.P. Lovecraft in his 1936 novella “At the Mountains of Madness.” In Lovecraft’s telling, Shoggoths were massive, blob-like monsters made out of iridescent black goo, covered in tentacles and eyes.

Shoggoths landed in the A.I. world in December, a month after ChatGPT’s release, when Twitter user @TetraspaceWest replied to a tweet about GPT-3 (an OpenAI language model that was ChatGPT’s predecessor) with an image of two hand-drawn Shoggoths — the first labeled “GPT-3” and the second labeled “GPT-3 + RLHF.” The second Shoggoth had, perched on one of its tentacles, a smiley-face mask.

     pic.twitter.com/wRVhjjSt0Y

    — tetraspace 💎 (@TetraspaceWest) December 30, 2022

In a nutshell, the joke was that in order to prevent A.I. language models from behaving in scary and dangerous ways, A.I. companies have had to train them to act polite and harmless. One popular way to do this is called “reinforcement learning from human feedback,” or R.L.H.F., a process that involves asking humans to score chatbot responses, and feeding those scores back into the A.I. model.

Most A.I. researchers agree that models trained using R.L.H.F. are better behaved than models without it. But some argue that fine-tuning a language model this way doesn’t actually make the underlying model less weird and inscrutable. In their view, it’s just a flimsy, friendly mask that obscures the mysterious beast underneath.

@TetraspaceWest, the meme’s creator, told me in a Twitter message that the Shoggoth “represents something that thinks in a way that humans don’t understand and that’s totally different from the way that humans think.”

Comparing an A.I. language model to a Shoggoth, @TetraspaceWest said, wasn’t necessarily implying that it was evil or sentient, just that its true nature might be unknowable.

“I was also thinking about how Lovecraft’s most powerful entities are dangerous — not because they don’t like humans, but because they’re indifferent and their priorities are totally alien to us and don’t involve humans, which is what I think will be true about possible future powerful A.I.”

The Shoggoth image caught on, as A.I. chatbots grew popular and users began to notice that some of them seemed to be doing strange, inexplicable things their creators hadn’t intended. In February, when Bing’s chatbot became unhinged and tried to break up my marriage, an A.I. researcher I know congratulated me on “glimpsing the Shoggoth.” A fellow A.I. journalist joked that when it came to fine-tuning Bing, Microsoft had forgotten to put on its smiley-face mask.

Got a pack of LLM shoggoth stickers!! I ordered extra, so if anybody wants one( and lives on the east coast) DM me. First come first serve pic.twitter.com/UhmhCDyacR

— MoreWrong (@wrong_more) March 17, 2023

Eventually, A.I. enthusiasts extended the metaphor. In February, Twitter user @anthrupad created a version of a Shoggoth that had, in addition to a smiley-face labeled “R.L.H.F.,” a more humanlike face labeled “supervised fine-tuning.” (You practically need a computer science degree to get the joke, but it’s a riff on the difference between general A.I. language models and more specialized applications like chatbots.)

Today, if you hear mentions of the Shoggoth in the A.I. community, it may be a wink at the strangeness of these systems — the black-box nature of their processes, the way they seem to defy human logic. Or maybe it’s an in-joke, visual shorthand for powerful A.I. systems that seem suspiciously nice. If it’s an A.I. safety researcher talking about the Shoggoth, maybe that person is passionate about preventing A.I. systems from displaying their true, Shoggoth-like nature.

In any case, the Shoggoth is a potent metaphor that encapsulates one of the most bizarre facts about the A.I. world, which is that many of the people working on this technology are somewhat mystified by their own creations. They don’t fully understand the inner workings of A.I.

language models, how they acquire new capabilities or why they behave unpredictably at times. They aren’t totally sure if A.I. is going to be net-good or net-bad for the world. And some of them have gotten to play around with the versions of this technology that haven’t yet been sanitized for public consumption — the real, unmasked Shoggoths.

That some A.I. insiders refer to their creations as Lovecraftian horrors, even as a joke, is unusual by historical standards. (Put it this way: Fifteen years ago, Mark Zuckerberg wasn’t going around comparing Facebook to Cthulhu.)

And it reinforces the notion that what’s happening in A.I. today feels, to some of its participants, more like an act of summoning than a software development process. They are creating the blobby, alien Shoggoths, making them bigger and more powerful, and hoping that there are enough smiley faces to cover the scary parts.

Source

Desktop PCs in general, and gaming PC desktops in particular, can get rather cluttered inside with lots of cables. Today, at Computex 2023, ASUS showed off a graphics card prototype that doesn't need cables to provide it with power.

Instead, as PC Gamer reports, the custom ASUS NVIDIA GeForce 4070 card connects to a custom slot on a PC motherboard, that slot provides the card with up to 600W of power. PC Gamer stated:

Asus's hardware solution involves a secondary slot on the motherboard that looks similar in size and shape to a 1x PCIe connection. The graphics card part of the solution sticks out separate from and aft of the main 16x PCIe connector on the PCB.

PC Gamer says that ASUS's custom motherboard slot may actually be a passthrough for a 16-pin connection to the actual power supply. That still means cables are used at some point to connect the PSU to the motherboard, but they don't actually connect directly to the graphics card.

Right now there's no word on if this new graphics card power configuration is limited to just ASUS cards and motherboards, or if it might also be adopted by other PC hardware makers. The article does say that these new motherboards will still work with graphics cards that need the old-fashioned power cable attached directly.

This kind of solution will not just help with making the inside of a PC desktop a bit cleaner. It will also allow people who build or upgrade their own PCs a simpler way to switch out graphics cards with new GPUs.

Right now, there's no word on when, or even if, ASUS will launch motherboards and graphics cards with this new slot-based power solution. If and when that does happen, we expect that it will cost a bit more to buy compared to standard cards and motherboards.

ASUS shows off a graphics card that uses a custom motherboard slot for its power

During the first quarter of this year, Canalys reported that notebooks accounted for 11.4 million units and desktops accounted for 2.7 million units. The notebook shipments were down 31% year-over-year while desktops were down 28%.

The company also looked at tablets and found that these held up better than PCs with just a decline of shipments by 7% down to 10.8 million units. This segment was greatly helped by sales of Apple’s iPad which saw the most shipments in the last two quarters than any time since Q1 2021.

“A confluence of factors led to the US PC market bottoming out at the start of 2023,” said Ishan Dutt, Principal Analyst at Canalys. “The expected seasonal drop after the holidays coincided with the channel finalizing its inventory correction, resulting in muted sell-in of PCs.“

“On the commercial front, businesses continue to grapple with economic pressures, including three further interest rate hikes since the start of the year. Budget-conscious organizations have extended the life cycle of their existing PCs, with many opting only for critical upgrades.”

As mentioned, the iPad has seen increasing shipments recently. In the first quarter of 2023, it saw 5.4 million shipments compared to 4.4 million in the same period the year before. That’s a 20.5% growth.

The only other tablet company to see growth was TCL but the growth was a measly 0.9%, nothing compared to the growth Apple saw. Amazon, Samsung, and Microsoft all saw tablet shipments fall - Microsoft was worst off seeing a decline of 30.4%.

In the PC section, no company saw any annual growth. Ironically, the company hit the hardest in the PC market was Apple which saw an annual growth of -35.8%. Lenovo was not too far behind at -35%. The company that held up best was HP which saw growth of -13.2%.

With increasing interest rates and the powerhouse of Europe, Germany, going into recession, it won’t be too surprising if we see much of the developed world fall into recession too. This slowdown could knock off Canalys prediction, but we’ll have to wait and see.

US PC market may have bottomed out, recovery predicted in Q4

The updated Windows Release Health dashboard contains all the details about the issue that crippled cameras in ARM-based Windows PCs. According to Microsoft, the bug affects devices with Qualcomm 8cx Gen 1, Gen 2, Microsoft SQ1, and Microsoft SQ2 processors running Windows 11 version 22H2 and 21H2, and Windows 10 22H2. In other words, it is not Surface-exclusive. At the same time, external USB cameras continue operating without issues.

A few days ago, users started reporting that integrated cameras work again, thanks to Microsoft deploying a "critical troubleshooter" to mitigate the issue on most affected Windows devices. That troubleshooter kicked in automatically and did not require any action from the user. You can check your ARM-powered PC by looking for the following entry in the troubleshooting history:

Hardware and devices troubleshooter

Automatically change system settings to fix a problem on your device

Although cameras are working again, Microsoft says the mitigation is temporary. Moreover, the fix is not available on managed devices, with troubleshooters disabled by administrators. Resolving the camera bug on those PCs requires executing the following command with administrator privileges:

reg add "HKLM\SYSTEM\CurrentControlSet\Control\Qualcomm\Camera" /v EnableQCOMFD /t REG_DWORD /d 0 /f

You can use the Command Line method on a non-managed device if it has not yet received the critical troubleshooter. However, Microsoft warns that it is intended for the affected devices only, so do not mess with your system's registry if the integrated cameras are working.

Another thing worth noting is that the workaround has a side effect: Microsoft says it might disable some camera features or lower the image quality. The company is working with OEMs on a new camera driver to bring cameras back to their normal behavior.

Microsoft temporarily fixed broken webcams on ARM PCs by lowering their image quality

This upcoming processor promises to enhance the AI capabilities of laptops and PCs, empowering users with improved performance and energy efficiency. In this article, we delve into the key features and potential impact of Intel's Meteor Lake processor in driving the AI revolution.

Intel Meteor Lake will increase the accessibility of local AI technologies - Image courtesy of Intel Corporation

Intel Meteor Lake will power PCs for AI

Scheduled to be launched later this year, Intel Meteor Lake boasts circuitry specifically optimized to accelerate various AI tasks. By offloading AI processing to the processor itself, this innovation ensures that AI workloads no longer solely rely on cloud-based computing.

One notable application is the enhancement of AI algorithms for videoconferencing, enabling features such as better background blurring or replacement. John Rayfield, leader of Intel's client AI work, highlights the potential for laptops to actively contribute to the AI revolution, bringing improved privacy, security, and responsiveness by reducing network delays.

The shift toward local AI processing

While the extent to which AI workloads will transition from the cloud to local devices remains uncertain, there are indications of a growing trend. Applications like Adobe Photoshop and Lightroom extensively employ AI for tasks such as image recognition and editing.

Moreover, technology giants like Microsoft are integrating AI chatbots directly into their operating systems. The inclusion of AI-specific accelerators within processors, as seen in smartphone processors and Apple's M-series Mac processors, may further incentivize developers to create AI-driven software.

Intel Meteor Lake might allow Intel to hold tight on the ongoing AI craze

Meteor Lake and GPU synergy

Graphical processing units (GPUs) have already demonstrated their prowess in accelerating AI tasks. Meteor Lake complements GPUs by introducing a specialized AI accelerator that excels in low-power scenarios. John Rayfield emphasizes that both the GPU and AI-specific accelerator can work together, delivering optimal AI performance.

While GPUs offer high-performance AI capabilities on PCs, the AI accelerator within Meteor Lake enhances efficiency, making it suitable for power-constrained environments.

Meteor Lake is significant for Intel's competitive strategy

Meteor Lake will be a significant milestone in Intel's strategy to regain its competitive standing. Addressing a critical weakness in comparison to Apple's M-series processors, it is designed for lower power operations while offering upgraded graphics acceleration, crucial for gaming and certain AI tasks.

Meteor Lake is also built on Intel 4, a new manufacturing process aimed at catching up with industry leaders like Taiwan Semiconductor Manufacturing Co. (TSMC) and Samsung. Intel's utilization of advanced manufacturing technology called Foveros allows for flexible and cost-effective stacking of multiple "chiplets" into a single, more powerful processor.

Only time will tell if Meteor Lake can take Intel to the front of the AI race.

Intel Meteor Lake will be the force behind AI's localization

PC Gamer got to see these prototypes up close and get more info on how they would work. One is called Dynamic Bimetallic Fins. The article states:

The bread, so to speak, is made up of two aluminum sheets, and the tasty filling is a copper sheet. The combined fins end up being around 1mm thick, which is roughly 3-4x thicker than your regular single fin, but the mixing of metals helps dissipate heat better—dropping temperatures by 3 degrees Celsius according to MSI.

Another possible solution is to use a thermoelectric cooler plate that would cover the GPU and its memory. The heat would then be removed to a radiator via liquid cooling. Yet another proposal is DynaVC, which would combine the card's vapor chamber and heat pipes. This is supposed to allow heat to move better between the GPU and the heat pipes. T

The most interesting, and most advanced, of MSI's graphics cooler prototypes is what the company is calling FushionChill. PC Gamer says:

Essentially, FushionChill is a liquid-cooled GPU with every part—including the radiator that usually comes connected via tubing on today's cards—within the graphics card shroud.

The design also includes a radiator that's deeper than normal ones and has an increased fin pitch. MSI claims that using this method can cool a GPU by as much as 10 degrees Celsius more than normal methods.

Since all of these concepts are in the prototype stage, there's no word one when these cooling solutions, if any, will be added to future MSI graphics card. It's possible a card could combine two of these methods for more effective cooling. However, they also have the potential to make graphics cards as a whole more expensive.

MSI shows off some prototypes of what could be used to cool your next PC graphics card

When many think of the term "liquid metal", they might picture in their head the T-1000 in Terminator 2: Judgement Day. However, the term is also used to describe a compound that can be used to cool PC CPUs. Today at Computex 2023, ASUS announced that it is using liquid metal for the first time inside one of its graphics cards.

ASUS's press release states that the new ROG Matrix GeForce RTX 4090 graphics card was made with a new and exclusive production process that will allow the liquid metal compound to be placed on the card's GPU and still allows the card to be installed vertically inside a PC. It adds:

By applying a UV resin above the surface-mounted capacitors that border the GPU die, and then adding two additional protective layers between the die and the main PCB, ASUS has ensured that the ROG Matrix GeForce RTX 4090 can leverage the incredible thermal transfer inherent to liquid metal with no risk of leakage.

In addition, the graphics card has a 360 mm radiator and three ROG MF-12S ARGB fans to keep the GPU and card cool. ASUS claims that this allows the card to have the "highest out-of-the-box boost clocks of any RTX 4090 in the world". However, the press release did not go into much detail on exactly what the clock speeds will be like on this card.

The design of the card and its housing is also something different from normal NVIDIA graphics cards. ASUS states:

The ROG Matrix RTX 4090 features an incredibly bold and modern design, with the GPU frame and housing built entirely out of metal. Sporting the first use of indirect ASUS Aura Lighting, the ROG Matrix GeForce RTX 4090 also boasts a 'hollow' chamber, with the main cold plate, circuit board and pump shroud protected by a larger aluminum frame. Bold ROG branding adorns the card, with lightbars in the central cavity and across the top of the card

So far, there's no word on when the ASUS ROG Matrix RTX 4090 will be released. There's also no indication of a price point, but considering all the extra hardware features, it's likely to be more expensive than normal NVIDIA GeForce 4090 cards.

ASUS is using liquid metal to help cool its latest GeForce RTX 4090 graphics card

Last week, Nvidia added close to $200 Billion in a day after the mid-week quarterly report for Q1 2023 where Jensen Huang, CEO of Nvidia noted that the company is reaping the benefits of being the first one to invest in various AI technologies.

Today, the company's stock price opened at $405 and hit a high of $419, making Nvidia the fourth company after Apple, Microsoft, Alphabet and Amazon to join the Trillion dollar club. As of writing the article,

Nvidia's market capitalization sat at $1.01 Trillion, up from $900 Billion last week. Nvidia has returned close to 30 percent in the last week, which is higher than the Nasdaq returns for the same period.

Over the last year, Nvidia has doubled investors' money. The investments in Artificial Intelligence (AI) made the company even more lucrative to investors.

The California-based chipmaker had a busy last couple of weeks where it doubled down on AI with announcements at Microsoft Build 2023 as well as at Computex 2023. These included ACE, a Chat GPT-like model for NPCs in games, a collaboration with WPP to introduce an innovative content engine, new AI enhancements for RTX GPUs, a new Ethernet switch focused on AI applications and a partnership with MediaTek to bring AI to cars.

While the company has been riding the AI wave, there has been skepticism surrounding the company's long-term growth and if it can sustain the current momentum. Nvidia saw similar jumps in stock prices during the crypto mining boom in 2020 which resulted in the GPU shortage. However, it didn't last long as the stock fell by more than 40% in 2022.

Nvidia becomes the first chipmaker to hit $1 Trillion market cap, thanks to the AI boom