Movers and Shakers in HPC: John Gustafson

By Caroline Connor

October 20, 2010

This is the first in a series of columns on movers and shakers in HPC, written by our newest contributing editor, Caroline Connor.

I had the pleasure of working with John Gustafson when he was Chief Technology Officer for ClearSpeed in 2007. Sure, I had heard about John, known for his work in HPC, describing the notion of weak scaling (Gustafson’s Law), introducing the first commercial computer cluster, winning the first Gordon Bell Award and all that. What surprised me was that there was so much more to John than the public persona. Here is a guy who is a former trampoline gymnast, built his own harpsichords at age 16, and grows orchids in his spare time. All of this is not lost upon me while I sit poolside at John’s lovely home, wondering what else I might uncover as I pull out my recorder.

HPCwire: John, you’re known for your “Reevaluating Amdahl’s Law” paper. Have you ever met Gene Amdahl? Is there any kind of debate still going on between the two of you?

John: (Laughs) I’ve met Gene, and have lunch with him every now and then; he lives right here in the Bay Area. We get along great. No, there’s no debate whatsoever. I’ve asked him things about his “law” that have been bothering me for years, and confirmed that he never meant his 1967 talk to be used to stop progress in parallel processing the way it has. He was debating Slotnick about the architecture of what would become the ILLIAC IV, saying that if you have one instruction stream, then the operating system part of that instruction stream will kill the parallelism. Gene told me that with modern systems, where every processor has its own instruction control, that argument doesn’t apply at all. So no, there’s no rivalry. I admire the man immensely and am honored to have any association with him.

HPCwire: So, what is behind your fascination with historical computers, like the 1939 Atanasoff-Berry Computer that you helped to reconstruct? It seems odd that a guy who works at the leading edge of supercomputing also works on machines that are a trillion times slower.

John: The technology of the era isn’t the important part; it’s what you do with it. So each generation rediscovers clever “tricks” about using tubes, discrete transistors, bit-slice logic, VLSI… and gives it a new name without realizing that there are many giants whose shoulders they could stand on. Another part of it is that Atanasoff has not received the credit he deserves for inventing electronic digital computing. Reconstructing his machine helped to set the record straight, and proved to people that his computer really worked.

HPCwire: I heard that you recently started managing Intel’s Ubiquitous High Performance Computing project for DARPA. What can you tell us about your new role and the project?

John: Well, this is my third time managing a grand “let’s build a big computer” project for DARPA. The first was when Steve Squires was leading the charge at DARPA in the 1980s, which led to the early hypercube projects and eventually to commodity clusters. The second was at Sun Microsystems when DARPA’s Bob Graybill was refocusing everyone on productivity instead of raw specs, and his HPCS program did a lot to realign people with the issues that really matter to computer users. Now, it’s the UHPC program. The goal is to produce an exaflop, or an exa-op, with less than 20 megawatts of electricity. If anyone can get the power efficiency that high in a general-purpose computer, it’s Intel. The aspect most interesting to me is the software part of the challenge. How much are we going to expose the architecture to the compiler developers and the library designers, versus the scientists and engineers who simply want to use the system to get work done? And do people have any idea about how power-hungry and numerically shaky our current “double precision” arithmetic will be when you’re doing a quintillion operations per second? I don’t think they do. So being able to direct such an effort is nothing less than fascinating. Finding time for outside activities just got a whole lot harder!

HPCwire: Speaking of which, what are your favorite hobbies, sports and other interests?

John: Oh, my. I didn’t expect that one. I was once a gymnast and pretty good on the trampoline, but that was quite a while ago. These days I spend my spare time playing piano and harpsichord. I actually learned to snow ski for the first time last year, and I plan on skiing more this season. Other than that, I usually enjoy the great California weather by swimming and hiking. At this point after taking on my new responsibilities for Intel, I feel lucky just to get outdoors enough to get some Vitamin D.

HPCwire: So, how old were you when you first started experimenting with electronics?

John: Oh my god, you would ask this. I don’t know whether to be embarrassed or proud about it, but I was six years old when I was assembling radio transmitters. I entered one in the science fair when I was in first grade, and won. What a geek I was! I saved up for a helium-neon laser and managed to get one when I was fifteen. I had indulgent parents who let me take over three rooms in the basement to make holograms, perform dubious chemical experiments, and generally do the kind of thing you might see in the Amateur Scientist column of Scientific American. By the time I entered Caltech as a freshman, I probably had about a thousand hours of hands-on lab experience, so the chemistry and physics courses seemed pretty easy.

My parents weren’t just indulgent, they were excellent guides. My mother had been an electronics technician at Collins Radio, now Collins-Rockwell, and my father was a chemical engineer turned MD, both as the result of World War II. One of my earliest memories was being taught about the polarity of batteries and electrolytic capacitors by my mother while trying to figure out what wasn’t working on the Heathkit breadboard circuit I just assembled. How geeky is that?

HPCwire: What are two or three interesting things about you that relatively few (or none) of your colleagues or friends know?

John: (Laughs) Well, my grandfather’s first cousin was Greta Garbo. Most of the family who came over from Sweden simply dropped the extra “s” in Gustafsson, but she probably followed someone’s good advice that even ‘Gustafson’ wouldn’t make it in Hollywood and changed her last name completely. Like my grandfather, she was from a poor farm on the outskirts of Stockholm.

Another thing people don’t know is that my father was the first guy to introduce computers into private hospitals in the US. People back then couldn’t figure out what possible use a computer could have in a hospital, but he persisted and said it could plan the diets of everyone, grade their psychological tests, maybe even monitor their electrocardiograms automatically. That was 1961 and 1962. When he visited IBM, I asked if I could go along. So here I was, this seven year old, touring one of the IBM sites in New York, slack-jawed at signs that said things like “Danger: Laser Light”… well, that was where they were working on the very first laser printers. I couldn’t understand why the reel-to-reel tape players kept starting and stopping; I thought they must all have been broken, and I wondered why no one could get them to work properly.

HPCwire: Just out of curiosity, why did you join ClearSpeed a few years ago? Based on your own personal experience, can you share any insights as to why some companies struggle in the HPC market place and so few survive?

John: Thomas Sterling told me once, “I figured out why you joined ClearSpeed: You’re re-living your youth.” I laughed, and knew exactly what he meant. I actually started my career at Floating Point Systems, a company that turned general-purpose computers into compute-intensive workhorses by adding special hardware for high FLOPS rates. I smiled when I got a pitch from ClearSpeed, who thought they’d invented the idea of using accelerators to plug into general-purpose boxes. I said, “So, your target markets are chemistry, structural analysis, and improving LINPACK scores, right?” To which they replied, amazed, “Yeah, how did you know that?” A few weeks later, I was offered the role of CTO and I agreed. It was a lot of fun while it lasted.

Seriously, in my personal opinion, HPC companies usually fail because they don’t identify their customers and their customer needs very accurately. Seymour Cray didn’t make that mistake; he was brilliant at knowing his customer base and what they wanted and needed.

HPCwire: I read just recently that Massively Parallel Technologies has announced a new software environment. As former CEO, can you share some of the history with us?

John: DARPA introduced me to MPT during the HPCS program, saying they had some very innovative ideas worth looking into. Gene Amdahl is on their technical advisory board, so I knew I should take them seriously. I was asked to take the reins to get them better connected to the mainstream HPC community, which I did. MPT has a technology for parallel programming that overlaps communication so well it allows scaling to millions of processors. The latest announcement is about something quite different. They’ve created a way to build programs that looks like the Apps Store, but hierarchical. Sort of the antithesis of open source; you get financial reward for every improvement you can make in a software supply chain. I would probably still be there had Intel not recruited me to direct their Santa Clara research lab in 2009. It was an offer I simply could not resist.

HPCwire: How would you describe yourself to someone who has never met you before, or knows nothing of your background?

John: Whew. That’s hard to do. I’d say that I’m an odd mixture of technophile and extrovert. I love public speaking, meeting people and talking to customers, which I notice isn’t true for a lot of scientist-engineer types. So I guess I’d say, “I’m a research scientist with a right brain.”

HPCwire: Lastly, what do you consider your greatest personal achievement?

John: Being influential in the adoption of parallel processing as a mainstream approach. Until 1988, when I wrote the paper about reevaluating Amdahl’s law, parallel processing was simply an academic curiosity that was viewed somewhat derisively by the big computer companies. When my team at Sandia — thank you, Gary Montry and Bob Benner — demonstrated that you could really get huge speedups on huge numbers of processors, it finally got people to change their minds. I am still amused by people out there gnashing their teeth about how to get performance out of multicore chips. Depending on what school they went to, they might think Amdahl proved that parallel processing will never work, or on the other hand, they might have read my paper and now have a different perception of how we use bigger computers to solve bigger problems, and not to solve the problems that fit existing computers. If that’s what I wind up being remembered for, I have no complaints.

About the Author

An avid HPC watcher and established technology marketing professional; Caroline resides in the California Bay Area and recently joined the HPCwire team as a contributing editor. You can reach her at [email protected].

Subscribe to HPCwire's Weekly Update!

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

HPE Announces New HPC Factory in Czech Republic

May 18, 2022

A week ahead of ISC High Performance 2022 (set to be held in Hamburg, Germany), supercomputing heavyweight HPE has announced a major investment in sovereign European computing: its first European factory, housed in the C Read more…

Hyperion Study Tracks Rise and Impact of Linux Supercomputers

May 17, 2022

That supercomputers produce impactful, lasting value is a basic tenet among the HPC community. To make the point more formally, Hyperion Research has issued a new report, The Economic and Societal Benefits of Linux Super Read more…

ECP Director Doug Kothe Named ORNL Associate Laboratory Director

May 16, 2022

The Department of Energy's Oak Ridge National Laboratory (ORNL) has selected Doug Kothe to be the next Associate Laboratory Director for its Computing and Computational Sciences Directorate (CCSD), HPCwire has learned. K Read more…

Google Cloud’s New TPU v4 ML Hub Packs 9 Exaflops of AI

May 16, 2022

Almost exactly a year ago, Google launched its Tensor Processing Unit (TPU) v4 chips at Google I/O 2021, promising twice the performance compared to the TPU v3. At the time, Google CEO Sundar Pichai said that Google’s datacenters would “soon have dozens of TPU v4 Pods, many of which will be... Read more…

Q&A with Candace Culhane, SC22 General Chair and an HPCwire Person to Watch in 2022

May 14, 2022

HPCwire is pleased to present our interview with SC22 General Chair Candace Culhane, program/project director at Los Alamos National Lab and an HPCwire 2022 Person to Watch. In this exclusive Q&A, Culhane covers her Read more…

AWS Solution Channel

shutterstock 1103121086

Encoding workflow dependencies in AWS Batch

Most users of HPC or Batch systems need to analyze data with multiple operations to get meaningful results. That’s really driven by the nature of scientific research or engineering processes – it’s rare that a single task generates the insight you need. Read more…

Argonne Supercomputer Advances Energy Storage Research

May 13, 2022

The lack of large-scale energy storage bottlenecks many sources of renewable energy, such as sunlight-reliant solar power and unpredictable wind power. Researchers from Lawrence Livermore National Laboratory (LLNL) are w Read more…

HPE Announces New HPC Factory in Czech Republic

May 18, 2022

A week ahead of ISC High Performance 2022 (set to be held in Hamburg, Germany), supercomputing heavyweight HPE has announced a major investment in sovereign Eur Read more…

Google Cloud’s New TPU v4 ML Hub Packs 9 Exaflops of AI

May 16, 2022

Almost exactly a year ago, Google launched its Tensor Processing Unit (TPU) v4 chips at Google I/O 2021, promising twice the performance compared to the TPU v3. At the time, Google CEO Sundar Pichai said that Google’s datacenters would “soon have dozens of TPU v4 Pods, many of which will be... Read more…

Q&A with Candace Culhane, SC22 General Chair and an HPCwire Person to Watch in 2022

May 14, 2022

HPCwire is pleased to present our interview with SC22 General Chair Candace Culhane, program/project director at Los Alamos National Lab and an HPCwire 2022 Per Read more…

Supercomputing an Image of Our Galaxy’s Supermassive Black Hole

May 13, 2022

A supermassive black hole called Sagittarius A* (yes, the asterisk is part of it!) sits at the center of the Milky Way. Now, for the first time, we can see it. Read more…

Royalty-free stock illustration ID: 1919750255

Intel Says UCIe to Outpace PCIe in Speed Race

May 11, 2022

Intel has shared more details on a new interconnect that is the foundation of the company’s long-term plan for x86, Arm and RISC-V architectures to co-exist in a single chip package. The semiconductor company is taking a modular approach to chip design with the option for customers to cram computing blocks such as CPUs, GPUs and AI accelerators inside a single chip package. Read more…

Intel Extends IPU Roadmap Through 2026

May 10, 2022

Intel is extending its roadmap for infrastructure processors through 2026, the company said at its Vision conference being held in Grapevine, Texas. The company's IPUs (infrastructure processing units) are megachips that are designed to improve datacenter efficiency by offloading functions such as networking control, storage management and security that were traditionally... Read more…

Exascale Watch: Aurora Installation Underway, Now Open for Reservations

May 10, 2022

Installation has begun on the Aurora supercomputer, Rick Stevens (associate director of Argonne National Laboratory) revealed today during the Intel Vision event keynote taking place in Dallas, Texas, and online. Joining Intel exec Raja Koduri on stage, Stevens confirmed that the Aurora build is underway – a major development for a system that is projected to deliver more... Read more…

Intel’s Habana Labs Unveils Gaudi2, Greco AI Processors

May 10, 2022

At the hybrid Intel Vision event today, Intel’s Habana Labs team launched two major new products: Gaudi2, the second generation of the Gaudi deep learning training processor; and Greco, the successor to the Goya deep learning inference processor. Intel says that the processors offer significant speedups relative to their predecessors and the... Read more…

Nvidia R&D Chief on How AI is Improving Chip Design

April 18, 2022

Getting a glimpse into Nvidia’s R&D has become a regular feature of the spring GTC conference with Bill Dally, chief scientist and senior vice president of research, providing an overview of Nvidia’s R&D organization and a few details on current priorities. This year, Dally focused mostly on AI tools that Nvidia is both developing and using in-house to improve... Read more…

Royalty-free stock illustration ID: 1919750255

Intel Says UCIe to Outpace PCIe in Speed Race

May 11, 2022

Intel has shared more details on a new interconnect that is the foundation of the company’s long-term plan for x86, Arm and RISC-V architectures to co-exist in a single chip package. The semiconductor company is taking a modular approach to chip design with the option for customers to cram computing blocks such as CPUs, GPUs and AI accelerators inside a single chip package. Read more…

Facebook Parent Meta’s New AI Supercomputer Will Be ‘World’s Fastest’

January 24, 2022

Fresh off its rebrand last October, Meta (née Facebook) is putting muscle behind its vision of a metaversal future with a massive new AI supercomputer called the AI Research SuperCluster (RSC). Meta says that RSC will be used to help build new AI models, develop augmented reality tools, seamlessly analyze multimedia data and more. The supercomputer’s... Read more…

AMD/Xilinx Takes Aim at Nvidia with Improved VCK5000 Inferencing Card

March 8, 2022

AMD/Xilinx has released an improved version of its VCK5000 AI inferencing card along with a series of competitive benchmarks aimed directly at Nvidia’s GPU line. AMD says the new VCK5000 has 3x better performance than earlier versions and delivers 2x TCO over Nvidia T4. AMD also showed favorable benchmarks against several Nvidia GPUs, claiming its VCK5000 achieved... Read more…

In Partnership with IBM, Canada to Get Its First Universal Quantum Computer

February 3, 2022

IBM today announced it will deploy its first quantum computer in Canada, putting Canada on a short list of countries that will have access to an IBM Quantum Sys Read more…

Supercomputer Simulations Show How Paxlovid, Pfizer’s Covid Antiviral, Works

February 3, 2022

Just about a month ago, Pfizer scored its second huge win of the pandemic when the U.S. Food and Drug Administration issued another emergency use authorization Read more…

Nvidia Launches Hopper H100 GPU, New DGXs and Grace Superchips

March 22, 2022

The battle for datacenter dominance keeps getting hotter. Today, Nvidia kicked off its spring GTC event with new silicon, new software and a new supercomputer. Speaking from a virtual environment in the Nvidia Omniverse 3D collaboration and simulation platform, CEO Jensen Huang introduced the new Hopper GPU architecture and the H100 GPU... Read more…

PsiQuantum’s Path to 1 Million Qubits

April 21, 2022

PsiQuantum, founded in 2016 by four researchers with roots at Bristol University, Stanford University, and York University, is one of a few quantum computing startups that’s kept a moderately low PR profile. (That’s if you disregard the roughly $700 million in funding it has attracted.) The main reason is PsiQuantum has eschewed the clamorous public chase for... Read more…

Leading Solution Providers

Contributors

Nvidia Dominates MLPerf Inference, Qualcomm also Shines, Where’s Everybody Else?

April 6, 2022

MLCommons today released its latest MLPerf inferencing results, with another strong showing by Nvidia accelerators inside a diverse array of systems. Roughly fo Read more…

D-Wave to Go Public with SPAC Deal; Expects ~$1.6B Market Valuation

February 8, 2022

Quantum computing pioneer D-Wave today announced plans to go public via a SPAC (special purpose acquisition company) mechanism. D-Wave will merge with DPCM Capital in a transaction expected to produce $340 million in cash and result in a roughly $1.6 billion initial market valuation. The deal is expected to be completed in the second quarter of 2022 and the new company will be traded on the New York Stock... Read more…

Intel Announces Falcon Shores CPU-GPU Combo Architecture for 2024

February 18, 2022

Intel held its 2022 investor meeting yesterday, covering everything from the imminent Sapphire Rapids CPUs to the hotly anticipated (and delayed) Ponte Vecchio GPUs. But somewhat buried in its summary of the meeting was a new namedrop: “Falcon Shores,” described as “a new architecture that will bring x86 and Xe GPU together into a single socket.” The reveal was... Read more…

Industry Consortium Forms to Drive UCIe Chiplet Interconnect Standard

March 2, 2022

A new industry consortium aims to establish a die-to-die interconnect standard – Universal Chiplet Interconnect Express (UCIe) – in support of an open chipl Read more…

Julia Update: Adoption Keeps Climbing; Is It a Python Challenger?

January 13, 2021

The rapid adoption of Julia, the open source, high level programing language with roots at MIT, shows no sign of slowing according to data from Julialang.org. I Read more…

Nvidia Acquires Software-Defined Storage Provider Excelero

March 7, 2022

Nvidia has announced that it has acquired Excelero. The high-performance block storage provider, founded in 2014, will have its technology integrated into Nvidia’s enterprise software stack. Nvidia is not disclosing the value of the deal. Excelero’s core product, Excelero NVMesh, offers software-defined block storage via networked NVMe SSDs. NVMesh operates through... Read more…

India Launches Petascale ‘PARAM Ganga’ Supercomputer

March 8, 2022

Just a couple of weeks ago, the Indian government promised that it had five HPC systems in the final stages of installation and would launch nine new supercomputers this year. Now, it appears to be making good on that promise: the country’s National Supercomputing Mission (NSM) has announced the deployment of “PARAM Ganga” petascale supercomputer at Indian Institute of Technology (IIT)... Read more…

Google Launches TPU v4 AI Chips

May 20, 2021

Google CEO Sundar Pichai spoke for only one minute and 42 seconds about the company’s latest TPU v4 Tensor Processing Units during his keynote at the Google I Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire