NVIDIA moving up and beyond China chip sanctions

Founder Jen-Hsun Huang donating big to America’s supercomputer future while finding a quick work-around to keep selling CPUs to China

NVIDIA CEO Jen-Hsun “Jensen” Huang and his wife Lori’s US$50 million donation for a new supercomputer research center at their alma mater Oregon State University (OSU) should remind investors, analysts and political economists that there is life after sanctions on China for America’s chipmakers.

Announced on October 14 and scheduled to open in 2025, the Jen-Hsun and Lori Huang Collaborative Innovation Complex will feature a supercomputer incorporating NVIDIA’s most advanced central and graphics processing units (CPUs and GPUs), a clean room, laboratories, an “extended reality” theater and other facilities.

The supercomputer is expected to be one of the fastest available for university research, capable of handling the most complex artificial intelligence (AI) models and digital twin simulations.

Scott Ashford, dean of the College of Engineering, said on the university’s website that the facility “…will help OSU be recognized as one of the world’s leading universities for artificial intelligence and robotics. It will transform not only the College of Engineering, but the entire university, and have an economic and environmental impact on the state of Oregon and the nation.”

Jen-Hsun and Lori Huang met while studying at OSU’s College of Engineering. They issued a statement saying: “We discovered our love for computer science and engineering at OSU. We hope this gift will help inspire future generations of students also to fall in love with technology and its capacity to change the world.”

“AI is the most transformative technology of our time. To harness this force, engineering students need access to a supercomputer, a time machine, to accelerate their research. This new AI supercomputer will enable OSU students and researchers to make very important advances in climate science, oceanography, materials science, robotics and other fields.”

OSU provost Edward Feser stated that “The collaborative innovation complex will be a key component of efforts championed by federal and state, business and academic leaders to support the competitiveness of Oregon’s semiconductor industry.”

Oregon Senator Ron Wyden said, “It’s no secret that advanced computer chips are the linchpin of the 21st-century economy. This state-of-the-art facility provides opportunity for Oregon State faculty and students to make generation-defining discoveries to push our tech industry forward.”

The total cost of the supercomputer complex is estimated at $200 million. OSU has already raised another $50 million in addition to the Huangs’ donation, is in the process of raising an additional $25 million and will ask the Oregon state legislature for the remaining $75 million.

Oregon State University’s new supercomputer complex will be running towards the AI future. Image: Agility Robotics

Oregon State is one of three land-, sea-, space- and sun-grant universities in the US. As such, it receives federal benefits related to agriculture, fisheries and marine science, space science and engineering, and bio-energy technologies.

Founded in 1889, OSU’s College of Engineering offers coursework on topics raning from bridges and dams to computer science and artificial intelligence. Alumni innovations, according to its website, include the first replacement heart valve, the computer mouse and the concept of email. 

Spin-offs from OSU include small modular nuclear reactor developer NuScale and advanced photoresist maker Inpria. NuScale works with governments and engineering companies in North America, Europe, Japan and South Korea. Inpria was recently purchased by Japanese photoresist maker JSR.

Jen-Hsun Huang was born Taiwan, from where his family moved to the US when he was nine years old. After receiving his undergraduate degree from OSU and a master’s degree from Stanford, both in electrical engineering, he worked for LSI Logic and Advanced Micro Devices. Huang founded Nvidia in 1993 together with Chris Malachowsky and Curtis Priem, who had previously worked at Sun Microsystems.

In 2020, Malachowsky and NVIDIA each donated $25 million to the University of Florida’s HiPerGator AI supercomputer project.

In 2021, Huang was awarded the Semiconductor Industry Association’s Robert N Noyce Award – its highest award, named for the co-founder of Fairchild Semiconductor and Intel – and the Institute of Electrical and Electronics Engineers (IEEE) Founder’s Medal in 2020. In 2019, he was named the world’s best-performing CEO by the Harvard Business Review.

NVIDIA introduced its first GPU in 1999, which was used in computer game machines (the 1994 PlayStation featured a “Sony GPU” designed by Toshiba). GPUs have become key components in numerous high-tech applications including robotics, autonomous vehicles and, most prominently, AI.

On September 1, the US Commerce department ordered NVIDIA to stop exporting its top-of-the-line A100 and H100 GPUs to China. Nvidia’s share price promptly dropped 12% and by mid-October, it had declined by almost 70% from its 52-week high. But most other semiconductor stocks were also down due to cyclical weakness in the industry and NVIDEA’s share price has since rebounded by more than 40%.  

The China-related sanctions were expected to cost NVIDIA about $400 million in lost sales, but on November 7, the company announced that it had started production of a scaled-down GPU, the A800, specifically for AI applications in the Chinese market.

NVIDIA is pivoting quickly from US Commerce Department export sanctions on China. Image: NVIDIA blog site

According to a company statement, the processor “meets the US government’s clear test for reduced export control and cannot be programmed to exceed it.” Demand for the new device is reportedly high.

The $400 million sanctions hit represents less than 7% of Nvidia’s third-quarter revenue, which was down 17% year on year to $5.93 billion. While a significant blow to business, it is already proving less-than crippling and at least partially offset.

Speaking at the NVIDIA GPU Technology Conference in September, Huang said

“Moore’s Law’s dead… the ability for Moore’s Law to deliver twice the performance at the same cost, or at the same performance, half the cost, every year and a half, is over. It’s completely over, and so the idea that a chip is going to go down in cost over time, unfortunately, is a story of the past.”

But, he added, “Computing is advancing at incredible speeds. The engine propelling this rocket is accelerated computing and its fuel is AI.”

This explains why the US government is obsessed with stifling China’s AI development and why Nvidia will have plenty of work regardless of the Commerce Department’s strict new export restrictions.

NVIDIA is a fabless design company, meaning it has no factories. TSMC currently makes its most advanced integrated circuits (ICs) in Taiwan. It has also used Samsung’s foundry services and in the future may outsource to Intel.

Huang told the press in March that Intel is “interested in us using their foundries. We’re very interested in exploring it.” In April, Intel announced an expansion of its factory in Hillsboro, Oregon, where, perhaps tellingly, its primary R&D facilities are also located.

Source