HOW IS CHINA USING U.S. SUPERCOMPUTERS?

September 8, 2000

FEATURES & COMMENTARY

New York, N.Y. — David Ruppe reports that the Clinton administration in the past three years has approved for export to China hundreds of high-performance computers capable of aiding that country’s nuclear and other weapons programs. And while U.S. law requires the government to check up on the machines to make sure they are not used to improve China’s military might, the Commerce Department, which is supposed to perform the checks, has conducted only a handful.

The latest publicly available statistics, for instance, show that 190 U.S. high performance computers, capable of performing 2 billion operations per second (or 2,000 MTOPS) or higher, were exported to China in fiscal year 1998 and are required, by law, to have such on site “end-use” checks. But of those 190 HPCs, Commerce investigators completed a check on only one, according to a recent congressional report.

As many as 600 computers above 2,000 MTOPS were shipped by U.S. companies to China between 1996 – when the Clinton administration greatly relaxed export controls on the machines – and the end of 1998. But reports show that as few as three of those were ever checked up on by Commerce. “We have not done all of the ones that we want to,” Commerce Undersecretary Bill Reinsch, whose Bureau of Export Administration would perform the checks, told ABCNEWS.com in a recent interview.

The problem has repeatedly raised concerns in Congress. “Given the lack of a proven and effective verification regime, it is possible that [certain high-performance computers] have been diverted for unauthorized uses,” said a report last spring by a bipartisan congressional committee specially tasked to assess the impact of U.S. exports to China. The committee, led by Rep. Christopher Cox, R-Calif., warned that China could be using U.S. supercomputers to upgrade and maintain nuclear and chemical weapons, better equip its forces with aircraft and submarines, develop a reliable and accurate ballistic and cruise missile force, and improve its computer warfare, anti-submarine warfare and communications capabilities.

The main reason for the lack of checks, according to Reinsch, is that the Chinese government rejects requests to visit most of the computers, citing national sovereignty. Under procedures established in June 1998, the Commerce Department must ask permission of its Chinese counterpart before U.S. inspectors can perform the checks. “They don’t say ëyes’ to all of them,” says Reinsch.

The checks involve sending teams of Commerce investigators, usually two or three people, to the various sites in a country where the computers are supposed to be housed, such as at banks, academic institutions and telecommunications centers. The investigators seek to verify, among other things, that the machine is where it should be, has not been upgraded and is being used for the purpose Commerce approved it for, and not for military purposes, such as to design, manufacture or test of nuclear, chemical, or biological weapons. Those checks the Chinese have allowed did not turn up any evidence of improper use, Reinsch said. But Reinsch acknowledged the Chinese may only be allowing checks on computers involved in approved purposes – and not those performing forbidden operations. “By asking [to visit] all of them, we effectively allow them to pick and chose which ones they say yes to. And I don’t have much doubt they’re saying yes to the ones that are benign,” said Reinsch.

The checks are required by a 1997 law for all computers over 2,000 MTOPS exported to China and other “Tier 3” countries identified by the government as security or technology proliferation concerns. But the law does not include a time limit for conducting the checks – and as a result, Commerce says it has not violated the law by failing to perform checks on most of the computers. And since China says it will not allow checks more than six months after the equipment arrives, it seems unlikely most of the previously exported computers will ever be checked on. Even if the checks were allowed, Reinsch says, Commerce doesn’t have the funds to send investigators to check on the hundreds of high performance computers exported to China. And even if U.S. inspectors were given access to computers designing weapons for the military, experts say, it would be difficult to determine how the computers are used when the inspectors aren’t there.

Meanwhile, the Clinton administration continues to allow high performance computer exports to China and is planning a new regulation change in January, allowing even more powerful computers to go to Chinese civilian and military buyers.

China began buying American supercomputers after the Clinton administration significantly loosened export controls on the technology in 1996. Before that, experts say, the country had no high performance computers. Since then, U.S. sales of computers between 2,000 and 7,000 MTOPS to the China have soared. While only 23 were exported in 1996, 123 were allowed in 1997, and as many as 434 were allowed for export during just the first three quarters of 1998.

The new regulations, which take effect in January, would raise the ceiling at which computer exports to Tier 3 countries would not require a Commerce Department license from 7,000 to 12,000 MTOPS for civilian recipients, and for from 2,000 to 7,000 MTOPS military recipients. “Here we are [about to] raise the threshold standard again, and the Chinese are basically thumbing their noses at us again. And what are we going to do about it? We’ll do nothing,” says Rep. Curt Weldon, R-Pa., a frequent critic of the administration’s technology export policies who served on the Cox Committee.

“Meanwhile, we’ve given China capabilities that they could never get on their own, to develop computerized weaponization programs, programs to design military technology, perhaps the capability to build a system – that can do nuclear test simulations. And it was all handed to them,” he said.

An Energy Department study released last year found that countries such as China, India and Pakistan could use 4,000 MTOPS computers to improve their nuclear weapon designs.

Today, desktop computers are sold with speeds above 2,000 MTOPS – and soon laptop computers just as fast could be available. High performance computers with strategic importance, however, tend to differ from very fast desktop computers in a number of ways. For instance: HPCs tend to have tens or hundreds of processors while desktop machines have one or two processors; most of the latest high performance computer’s are parallel processors, while most desk top machines are not; and HPCs have much more powerful memory access than desktop computers.

Such distinctions are not currently addressed by U.S. export law. So the adminstration’s planned relaxation of computer export restrictions should make sensitive high-performance computers, as well as personal computers, easier to obtain by entities in countries of concern.

Leading U.S. manufacturers of high performance computers include Silicon Graphics Inc., Sun Microsystems Inc., and Digital Equipment Corp., which was recently acquired by Compaq Computer Corp.

The U.S. government has been negotiating with Beijing for more than 15 years to allow end-use checks on high-tech equipment sold to China. Other countries – such as Israel, Russia and India – by and large have submitted to the checks. But China, which in recent years has bought many more American supercomputers than all other countries combined, has opposed the U.S. government checking up on how those computers are used, citing concern for national sovereignty.

In a June 1998 agreement touted by the Clinton administration, China appeared finally to consent to allow the checks. But the agreement, as it turned out, provided the Chinese broad discretion. According to one of the terms of the agreement, China could continue to refuse a check if it determined it would violate national sovereignty. Also, China could refuse a check if it determined it would cost too much to send Chinese officials to accompany the U.S. investigators, or if the computer was not shipped with a Chinese “end-user certificate,” documents the Chinese government said it would start issuing for supercomputer exports following the June 1998 agreement.

Checks were not performed for such reasons on 82 of the 190 supercomputers requiring checks in fiscal 1998, according to the Commerce Department. Checks on another 105 were not performed because the exports were made prior to the June 1998 agreement, after which China said it would start allowing the checks, the agency reports. Further, the agreement requires that checks be conducted by Chinese officials, not by U.S. representatives, and that checks cannot be performed six months after the machine arrives in China. China’s conditions on the checks, said the Cox committee report, “rendered the agreement useless.”

============================================================

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire