HPC-ETC: THE PAST MONTH IN HIGH PERF. COMPUTING

December 1, 2000

by Steven Witucki, Assistant Editor

SUPERCOMPUTING TAKES ANOTHER TURN

Once, supercomputing was the exclusive province of custom-designed machines with elaborate cooling systems made by companies like I.B.M. and Cray Research for government research laboratories and giant auto, petroleum and chemical companies. Then, starting in the early 1990’s, I.B.M. and new rivals like Hewlett-Packard, Sun Microsystems and Compaq Computer rapidly shifted to systems that lashed together the same processors used in common business computers, relying on software to break supercomputing’s giant challenges into bite-size pieces. Now the dominance of such clusters is being challenged by the arrival of start-up companies offering methods of doing supercomputing over the Internet on home and office computers that would otherwise be idle.

“There are 100 million machines hooked to the Internet, all of them doing nothing a lot of the time,” said James Gannon, chief technology officer of Parabon Computation, one of several first-time exhibitors in Dallas this month at supercomputing’s biggest trade show. “This kind of inefficiency just can’t exist in a free-market economy.” The potential of such Internet-based supercomputing, also known as grid computing, has been highlighted by the SETI@Home project (for Search for Extraterrestrial Intelligence), a nonprofit effort that has signed up two million computer owners to help crunch data obtained by radio telescopes scanning the skies for signs of life.

The long-term trend toward building supercomputing systems out of products developed for other sectors has clearly spread access to supercomputing, but some experts say the broadening of the market has diverted technology development away from the needs of hard-core users, who make up a $1 billion niche market, according to International Data. They argue that many problems detailed weather prediction, the behavior of the Internet, an atomic perspective on how proteins behave or what happens to the organs and other soft tissues of passengers in auto crashes need much more rapid data sharing than Internet grids or even dedicated clusters of off-the-shelf computers can generate. “We do the whole gamut of supercomputing, but the really tough problems get solved on real expensive hardware,” said Vincent F. Scarafino, manager of numerically intensive computing at Ford Motor. The big computer companies agree. Hence plans like that of I.B.M. to invest more than $100 million in Blue Gene, a supercomputer dedicated to studying genetics and other fundamental problems in biology, or the Japanese government’s commissioning of NEC to build the world’s most powerful supercomputer, called the Earth Simulator, to study global warming, El NiÒo cycles and other global environmental events.

ICANN GIVES SEVEN NEW DOMAIN NAMES

The push to expand the Internet’s naming system beyond the “.com” category came to a head this month as the technical administrators meet to select new suffixes such as .biz, .health, and .nom. But ICANN, the U.S. government-backed, nonprofit group charged with setting Internet naming conventions, faces a contentious week-long meeting here as backers of new names like .web to .kids and .xxx fight for last-minute acceptance. Controversy is nothing new to ICANN, the Internet Corp. for Assigned Names and Numbers, which has weathered criticism from all sides in the short-time since the U.S. government handed it control of the explosively growing communications medium. Many observers believe that ICANN’s own legitimacy hinges on the success of this expansion, the first broad policy move since the organization was formed at the White House’s invitation in 1998.

Out of 44 applications, the board of the Internet Corporation for Assigned Names and Numbers chose .biz, .info, .name, .pro, .museum, .aero. and .coop – and rejected a slew of other options hotly debated by industry players. The choice of the new domain names, which are expected to become available to businesses and consumers by the middle of next year, culminates a drawn-out process for setting the next stage of growth on the Internet. “This is a first giant step for domain-kind,” said Esther Dyson, chairman of ICANN. Major companies involved in winning bids to operate the huge databases holding Web site addresses, also called registries, included VeriSign Inc., which currently enjoys a near-monopoly as the sole registry operator for all domain names not ending in a country suffix, International Business Machines Corp. and Register.com, a fast-growing U.S. reseller of Web site addresses. Some of the new domain names, like .info and .name, will be open to almost anyone to register starting by the middle of next year. Others, such as .museum, and .biz, will be restricted to members of companies or relevant fields. One of the most vied-for domain names, .web, was at the last moment taken out of the “approved” basket because of a controversy over the ownership. Affilias, a consortium of big companies including VeriSign, was wrangling with a small California company, Image Online Design Inc., which said it registered tens of thousands of users for .web in the past several years due to what it claims was a prior agreement. An impassioned speech by Vint Cerf, an ICANN board member who was later elected new board chairman, on behalf of Image Online, helped sway the board at the last moment, which granted Affilias the .info domain name, considered to be a less popular domain name, instead.

RAYTHEON TO UPGRADE SUPERCOMPUTER LABORATORY

Raytheon Company has been awarded a $34 million contract to upgrade the Geophysical Fluid Dynamics Laboratory (GFDL) supercomputer laboratory in Princeton, N.J. The upgrade is expected to improve the nation’s climate prediction and weather forecasting capabilities. GFDL is a federal research laboratory in the Office of Oceanic and Atmospheric Research, part of the U.S. Department of Commerce’s National Oceanic and Atmospheric Administration (NOAA). The laboratory performs comprehensive, long lead-time research that is fundamental to the mission of NOAA. One of the goals of this research is to expand the scientific understanding of the physical processes governing the behavior of the atmosphere and the oceans as complex fluid systems. These systems can be modeled mathematically and their phenomenology can be studied by complex computer simulations. Using this information can improve climate prediction and weather forecasting. GFDL is one of the nation’s foremost computer laboratories concentrating primarily on the modeling of hurricanes and other large-scale weather phenomena. “The opportunity to be part of this GFDL team capitalizes on Raytheon’s capabilities in high performance computing and earth sciences,” said Ellis Bailey, Raytheon’s GFDL program manager. “The SGI Origin 3000 series technology delivered will greatly improve GFDL’s ability to meet the demands of its scientific research.”

Initial delivery consists of nine SGI Origin 3800 supercomputers each with 128 processors. The new computers will have the ability to perform more than 900 billion floating point arithmetic operations per second (900 Gigaflops) and will have more than four times the performance of the three Cray Research computers that they replace. SGI, a key subcontractor on the GFDL team, recently launched its new Origin 3000 series servers. Available immediately, the systems are based on the breakthrough SGI NUMAflex modular technology, and offer flexibility, resiliency, overall investment protection and superior performance, according to Jan Silverman, vice president, SGI Advanced Systems Marketing. The SGI NUMAflex modular technology is a “brick”-style system for constructing small-to-very large computer systems from a common set of building blocks. Financing for the project is being provided by SGI Solutions Finance. The system also includes more than 20 terabytes of high-speed disk storage that can transfer data in or out of the system at more than 10 gigabytes per second. Raytheon will also, as part of the contract, upgrade an existing automated tape archive storage system. Initial capacity of the tape archive will be approximately 500 terabytes.

QUANTUM DOTS MAY YIELD CHANGES IN COMPUTERS

Like Edison in his search to perfect the light bulb, if Supriyo Bandyopadhyay and his colleagues succeed, they will have a discovery that could revolutionize how the world works. Bandyopadhyay’s Quantum Device Laboratory at UNL is one of a handful of labs in the United States in the final round of two federal government funding competitions to develop quantum dot-based electronics. The Nebraska group has patented a technique that produces quantum dots, tiny structures that are 10,000 times smaller than the thickness of a human hair, but whose potential is staggering. In the next few decades, they could make binary computers obsolete and in a few years could make satellites safer from laser attacks. “There are many applications for quantum dots,” Bandyopadhyay said. “The obvious application is that you can make very small structures and if you can store information in them, you can have very high information storage density. You can also use these structures to do very efficient, very high-speed computation. You can build quantum computers.”

“As an example of what quantum computers can do, let’s say that you wanted to build a computer that has two to the 1,000th power (21,000) bits of data. You could never build a classical computer to do that because the number two to the 1,000th power is larger than the number of atoms in the known universe. With a quantum computer, all you would need is just 1,000 atoms to build a computer that powerful.” In other words, a device far too small to be seen by the naked eye would vastly outstrip in power and speed any computer now in existence. It would also do it without generating heat and may be capable of storing data indefinitely. In contrast, conventional computers produce large amounts of heat and have to refresh their data several times a second to avoid losing it. Don’t expect to run down to the local Radio Shack to buy one anytime soon, however. Bandyopadhyay and five fellow UNL electrical engineers studying nanotechnology, Rod Dillon, Ned Ianno, Latika Menon, Paul Snyder and Frazer Williams, have been working on quantum computer research for about three years and have succeeded in demonstrating new types of computer memory. But Bandyopadhyay said his team is probably five years away from being able to demonstrate a small-scale quantum computer in the lab, while commercial versions probably won’t be available for 20 to 25 years.

AC3 OPENS IN AUSTRALIA

The Australian Centre for Advanced Computing and Communications (ac3) opened its doors this month. Ac3 is part of a national group, The Australian Partnership for Advanced Computing (APAC), which has received $47.5 million in Federal and State government funding to set up supercomputer facilities across the country. The program got off to a bad start in October when APAC rejected the winning tenderer of its flagship supercomputer: the Sun Microsystems machine failed to meet key performance benchmarks. A re-tendering process is now due to close in December. Ac3, based at the Redfern Technology Park, is the only member of the group that has established itself as a privately run, fully commercial entity. However, it is 57 per cent owned by the NSW Government, which contributed $12 million in seed funding. NSW IT Minister Mr Kim Yeadon announced that the centre had opened with “three major business deals worth nearly $9 million”. The founding tenants are Superquant, Animated Biomedical Productions and SIRCA. However, the $9 million figure is dependent on two of the three companies gaining further government and venture capital funding, and even then appears ambitious compared to the companies’ own expectations.

SUPERCOMPUTER TO ANALYZE OBJECTS IN ORBIT

IBM has won a bid for a 320-processor supercomputer that will help the Air Force keep track of satellites, discarded rocket parts, space-suit gloves and other miscellany orbiting the Earth. The machine will run at the Maui High-Performance Computing Center, using its processing horsepower to refine otherwise blurry images of the 9,000 orbiting items spotted by Air Force radar and telescopes, Gene Bal, director of the center, said in an interview. The new computer will speed up the image processing so that Air Force telescopes can be set up to carefully examine objects before their orbit carries them below the horizon, he said. In addition, the new system, which cost an estimated $4 million to $5 million of a broader $10 million contract, will let the Air Force scrutinize more objects at once. “We apply numerical, mathematical algorithmic techniques to take an almost useless image of an object in space and convert that to a very good image they can use for identification,” Bal said.

Space junk is a major problem for those who consider launching space stations, satellites or space shuttles into orbit. The National Aeronautics and Space Administration said a ground crew lost contact with a British satellite called Cerise after an old fragment of an exploded rocket knocked of part of Cerise’s stabilization system. And an orbiting space shuttle docked with the Hubble Space Telescope was forced to hastily dodge another piece of rocket. A 1995 NASA study found that only 5 percent of orbiting objects were functioning spacecraft, much less than the 40 percent that are other manmade items such as dead spacecraft or hunks of discarded rocket engines. Keeping track of these objects is tough, particularly when they break into fragments, as two Russian Proton rocket parts did in the spring of 1999, one crumbling into 17 parts and another into 76. The machine will be used for more than just peering at space junk and newly launched spacecraft the United States wants to study. It also will help the U.S. Navy conduct simulated battles in the Pacific Ocean, Bal said. The IBM machine has 224GB of memory and nearly 3 terabytes of hard disk space, IBM said. The 320 processors are organized into 80 four-processor computers joined by a high-speed interconnect.

1,024-PROCESSOR SERVER INSTALLED IN NETHERLANDS

The Netherlands Organization for Scientific Research (NWO) and the Netherlands Computing Facilities (NCF) Foundation have implemented an SGI Origin 3000 series supercomputer from SGI at SARA Computing and Networking Services. The supercomputer will enable advanced fundamental and applied scientific research. The new facility was opened by HRH Prince Willem-Alexander. Mr. L.M.L.H.A Hermans, minister of education, culture and science, also attended. The SGI supercomputer will help the Dutch academic community to understand and resolve the world’s most complex scientific, technical and medical issues. Immediate areas of research will include climate research; computational medical science, water management and water quality calculations (areas of particular interest for the Prince); fluid dynamics and turbulence modeling; and computational chemistry, including drug design. The supercomputer, which is housed and operated by SARA, is one of the first European implementations of the new SGI Origin 3000 series and the first 1,024-processor SGI Origin 3000 series server to be installed worldwide. The SGI Origin 3000 series server is a highly scalable, high-performance SGI NUMA server with modular architecture, meaning each system can be tailored to exactly match performance and application requirements.

The new supercomputer features 1,024 MIPS processors that will deliver more than one TFLOPS (a trillion operations per second) of peak performance, 10TB (10,000GB) of online storage and 100TB of near-line StorageTek storage. Bob Bishop, chairman and CEO of SGI, said, “SGI was selected to manufacture, install and provide a six-year maintenance contract for the supercomputer in February of this year after extensive competitive evaluation by NCF. SGI consultants have worked closely with the Dutch research council and supercomputing authorities since the contract was awarded. We are delighted to provide them with a tailored solution capable of meeting their cutting edge scientific objectives. SGI has a strong reputation for meeting the needs of the most demanding customers in the high-performance computing sector.” “This facility will put one of the most advanced computer systems in the hands of Dutch scientists and engineers, allowing them to operate at the cutting edge of research,” said Dr. Reinder van Duinen, president of NWO.

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

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!

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…

Nvidia Appoints Andy Grant as EMEA Director of Supercomputing, Higher Education, and AI

March 22, 2024

Nvidia recently appointed Andy Grant as Director, Supercomputing, Higher Education, and AI for Europe, the Middle East, and Africa (EMEA). With over 25 years of high-performance computing (HPC) experience, Grant brings a 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…

Houston We Have a Solution: Addressing the HPC and Tech Talent Gap

March 15, 2024

Generations of Houstonian teachers, counselors, and parents have either worked in the aerospace industry or know people who do - the prospect of entering the fi 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