PETAFLOPS IN 2009: AN INTERVIEW WITH STEVE WALLACH

November 7, 2000

by Alan Beck, editor in chief LIVEwire

Dallas, Texas — SC2000’s keynote address was given by Steven J. Wallach. Wallach co-founded Convex Computer Corporation, along with Robert J. Paluck, former chairman and CEO, in 1982 and was the chief designer of the Convex C-Series, the world’s first affordable supercomputer, as well as the Exemplar Scalable Parallel Processor (SPP), HP/Convex.

Wallach is currently an advisor to CenterPoint ( http://www.centerpointvp.com ) Venture Partners, Dallas, Texas and Vice President of Technology of Chiaro Networks ( http://www.chiaro.com ), Richardson, Texas. He may be best known outside HPCN circles as the Data General engineer who was the principal architect of the 32-bit Eclipse MV superminicomputer series as described by Pulitzer Prize winner Tracy Kidder in The Soul of A New Machine.

Wallach holds 33 patents in various areas of computer design and held a joint appointment in the Graduate School of Management and Brown School of Engineering, Computer Science, Rice University for the 1998 and 1999 academic years. He is a member of the PITAC (Presidential Advisory Board on High Performance Computing, Communications, and Networking) and the advisory committee for the Hybrid Technology MultiThreaded Architecture (HTMT) a US DOD funded project to develop the concepts for a PETAFLOP computer. He is also a member of the National Academy of Engineering.

HPCwire interviewed Wallach to explore some of his current perspectives on the state of high performance computing:

HPCwire: Your SC2000 keynote is entitled “Petaflops in the Year 2009”. Is this realistic? What are the principal challenges HPC must meet to effect this goal?

WALLACH: This goal is more than realistic. One can make an argument that a petaflop computer system exists today, it is call the Web. It has been well documented how 1000’s of computers, distributed throughout the world, have been used to solve embarrassingly parallel applications. If we can apply 1,000,000 networked pc/workstations, we get a petaflop computer. Entropia is a example of an effort that is attempting to do this.

What my keynote address discusses is how to make a petaflop computer that is more general purpose (an oxymoron perhaps?), and that is located on one location (that also has to be re-examined). Much of the technology that is used and developed for GRID computing today will be used for the petaflop computer that I will describe.

The principal challenges have not really changed much in the last 10 years. We will need advances in: software; including compilers, OS, development environments, and the interconnect/memory system. Every time a new generation of processor is developed, with its own unique internal architecture, we stress the existing development and algorithmic environment.

We must also rethink the way we do storage. Petaflops of computing implies Petabytes of storage. I believe that architectures developed for web based and commercial storage systems will become the leading edge architectures for technical computing.

HPCwire: After pioneering supercomputing technology, you are now closely involved with both CenterPoint Venture Partners and Chiaro Networks. What do you hope to accomplish through these corporate efforts?

WALLACH: Well, I guess I am still an engineer at heart. I like to make things happen and like to ship product. The more disruptive the technology the happier I am. Today, that generally means doing things in a startup. So whether that means helping companies get started or directly getting involved in day to day operations. In fact, one can make an argument, that major companies throughout the world relay on startups for their new technology. As near as I can tell, all major technology companies have a venture capital group. These internal venture groups look for companies that have technologies that are strategic to the corporate mission. Intel Capital is perhaps the best example of these phenomena.

I recently gave some testimony before a US Senate Committee. This was to support upcoming NSF appropriations. One of the speakers, from the NSF, referred to one of their missions as “The Venture Capitalist of the first degree”. Meaning that government “invests” in research without a consideration of a financial return on investment, but a research return. I agree with this perspective.

When doing due diligence on companies seeking funding, it is fun to perform design reviews and/or make suggestions for improvement. Too many potential founders, try to impress venture capitalists with spread sheets, etc; in my book a spreadsheet is a random number generator. Also, with the CenterPoint and Sevin-Rosen funds, we have a keiretsu type of organization. In many cases, startups in the family help each other, when and where appropriate.

Personally, I am on the technical board of advisors of two startup companies; Chorum Technology (optical components) and Scale8 (Petabyte storage systems), and help out with some others.

HPCwire: As a member of the Presidential Advisory Board on High Performance Computing, Communications, and Networking you are in a unique position to observe the impact of policy and politics on HPC. How would you characterize your experiences in this arena? Are there frustrations and/or satisfactions that you find particularly noteworthy?

WALLACH: There are both; frustrations and satisfactions. The frustrations are the level of politics and what has to be “politically correct”. I will not go further, but Washington is Washington and politics is politics.

The satisfactions more then outweigh the frustrations. Helping our country by helping the members of the various branches of the government understand the importance of high performance computing. The one major recommendation of PITAC was that the US totally under spends for long-term basic research. Today; most of the research is for applied research. Long term basic research funding is needed to help solve the problems and develop the technologies that are needed 10 to 20 years from now. That is difficult to convince someone, who, perhaps only has a 4 to 6 year view. But we must increase funding levels for long-term basic research.

There are two aspects of high performance computing that are very important. One is for national security reasons. The ASCI program is a prime example of this. The other is the trickle down effect that high performance computing has on more commonplace applications. The extensive use of clusters and SMP’s for various web-based services would not have been made possible without the technology that was developed for high performance computing. Unfortunately this is not well understood nor appreciated.

HPCwire: When HPCwire interviewed you in 1997, you noted that knotty programming problems, often focused on algorithms and legacy code, were responsible for stymieing much Progress in HPC. Has this changed? How? Have architectures like Tera’s MTA changed the picture significantly?

WALLACH: No, not really. Legacy codes still prevail in the technical fields. The newest codes are web centric and are generally written in java. But they are rarely numerically intensive. Every time a new processor or system architecture is developed, the code generator and machine dependent optimizers have to be redone. And in many cases, application tuning is needed. I am convinced that this is becoming, if not already, an art and not a science. At Convex, I use to say that benchmarking and tuning a system is really a benchmarking and tuning test of your analyst.

Tera’s MTA is a significant advance in computer architecture. But to fully utilize its capabilities you still need to tune your algorithms and your code.

HPCwire: With the new century, a new generation of computer scientists is taking the reins of HPC development. What advice would you like to give them?

WALLACH: Try to start out with a clean sheet of paper. Also recognize that the biggest market for high performance computer and scalable parallel processors are the web centric servers. Applications like database, web hosting, storage for petabytes of data and media files, will dominate. And then incorporate numerical intensive features. The new generation also has to be more network and grid centric.

From a language perspective, we will continue to evolve FORTRAN, C, and JAVA. It appears that every 10 to 15 years or so, a new language is accepted, not by industry or government edict, but by a community ground swell. That is what happened with C and JAVA. So someone in the next 10 years or so will probably develop a new paradigm for software development that will be accepted. I have no idea what this will look like, but it will surely happen.

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

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