CSCS Benefits from HPC Challenge Benchmarks

By Nicole Hemsoth

October 7, 2005

HPCwire recently spoke with Marie-Christine Sawley, director of the Swiss National Supercomputing Center (CSCS), about its use of the relatively new HPC Challenge benchmark for recent large-scale supercomputer procurement. The HPC Challenge, co-sponsored by the National Science Foundation, U.S. Department of Energy and DARPA HPCS (High Productivity Computing Systems) program, was introduced at the SC2003 conference. Assembled by Jack Dongarra and Piotr Luszczek of the University of Tennessee, with collaborators from the U.S. and Europe, the benchmark suite includes Linpack and tests of other capabilities that can make a major difference in the real-world performance of HPC systems. Although HPCwire expects the U.S. government to begin using the HPC Challenge in procurements soon, CSCS may be the world's first major site to do this.

HPCwire: Tell me about CSCS.

Sawley: We are the Swiss National Supercomputing Center, located in Lugano, south of Zurich. We serve scientific researchers throughout Switzerland and beyond. CSCS was created in 1991 to provide leadership-class computing for high- end science nationwide and to support breakthrough science, academic partnerships and world-class connectivity. We are an autonomous unit of ETH- Zurich.

HPCwire: Why does your organization need a large-scale supercomputer?

Sawley: Since 2004, we've operated on a mandate to enrich our portfolio with some of the most demanding users and scientific applications. We had few of these applications at CSCS before, but we knew there were more out there. Users in this category were either trying to outsource their work on systems they could access, or they were limiting their work. We wanted to give them a way to scale up. We decided to extend our computing infrastructure with a tightly coupled, leadership-class system.

HPCwire: What were the goals for your “Horizon” procurement?

Sawley: We wanted a capability machine based on MPP architecture, with 1,000 or more CPUs and a fast interconnect. We ended up choosing a Cray XT3. Any Swiss researcher will be able to apply for computing time on this machine. The targeted applications domains are computational chemistry, materials science, molecular dynamics, physics and climate science. For the future, there could also be some interest in financial applications based on stochastic modeling.

HPCwire: How did you become aware of the HPC Challenge benchmark suite?

Sawley: We recognized the limitations of Linpack and witnessed the development of the HPC Challenge, especially through our partnership in SOS with Sandia and Oak Ridge.

HPCwire: Why did you decide to use HPC Challenge?

Sawley: Because we can measure and analyze the characteristic dimensions of a given supercomputer architecture with it. It allows you to base your investment not only on the sustained Gigaflop/s or wall clock time of a few applications, but to develop a procurement strategy which is part of an overall supercomputer portfolio management strategy. You also don't have to hand as many user codes over to the manufacturers for benchmarking and optimization during a call for tender project. When you have them run the HPC Challenge suite, they are running a suite of synthetic benchmarks that are well known in the community. This lowers the burden on the bidders, speeds your procurement project, and still allows you to estimate the effects of a given architecture on your key user applications by mapping their characteristic requirements onto the various HPCC benchmark test results.

HPCwire: How did the HPC Challenge fit into your other procurement criteria?

Sawley: The major challenge we faced was characterizing how applications would map onto architectures. Running HPC Challenge plus user codes turned out to be complementary. We assigned the heaviest weighting to the HPC Challenge results. Performance on user codes also played a role, of course, and we had to complete the procurement benchmark suite with another synthetic benchmark for file system I/O.

HPCwire: What was your experience using the HPC Challenge?

Sawley: Neither the bidders nor we had any major problems. It created a level playing field for the bidders and there was no resistance from them about using it, even though this was the first time they had to deal with the HPC Challenge in a procurement, at least in Europe. Most of the bidders approached it as a challenge, in the most positive sense. There were no surprises created by the HPC Challenge suite.

HPCwire: Since your pioneering use of the HPC Challenge, have you heard from other HPC sites that are interested in using it too?

Sawley: We have heard from a number of sites in the U.S. and a couple in Europe.

HPCwire: Would you use HPC Challenge again? Would you recommend it to others?

Sawley: We would definitely recommend it to others, and would certainly use it again for this kind of procurement, in which you are preparing for a major step forward. Our biggest machine before this procurement was 250 processors. If you want to take a big step forward, as we did, HPC Challenge is particularly well designed for high-level testing of your workload. There's always a certain amount of risk, of course. We had internal discussions on how to benchmark with codes that were in development, and using HPC Challenge seem to offer a sensible way to characterize the systems.

HPCwire: Do you think you'll succeed in attracting more demanding applications and users?

Sawley: Yes, because there's an unmet demand we will be able to address with our new Red Storm-based supercomputer. Some momentum has already started. Before this procurement, our largest applications used 128 CPUs; but as word got around that this bigger system was coming, ideas for scaling up really began flourishing. It has already spurred innovation.

HPCwire: What are your biggest challenges as an organization? How do you plan to overcome them?

Sawley: One challenge when you're growing is not to lose focus. We are mitigating this risk by adding scientific services for our users, including benchmarking, performance modeling, advanced visualization and other tools for data evaluation. So, we're enhancing our competencies. We need to deliver all this, of course. We have a performance agreement with our funding bodies, with key performance indicators we have to meet.

HPCwire: What is CSCS' position on industrial relations and technology transfer?

Sawley: Our belief is that simply selling cycles to industry is hard and you often don't learn much if the collaboration stops there. The most successful model, in our view and in our experience, is to construct projects with specific goals. Both parties learn in this model. With utility computing, you don't get that learning: the users take the added knowledge away with them. Technology transfer is complex.

At CSCS, our approach is to make sure we have a scientific leader, someone who acts as a principal investigator and can validate the science. Beyond that, innovation requires an ecosystem, an environment that fosters the open exchange of information and knowledge transfer. For our “Horizon” procurement and the progress that will come from it, Switzerland's Paul Scherrer Institut was the right partner, because they came to us with deep scientific knowledge and an unmet need. To run their HPC applications, they had to go outside Switzerland before. Their applications were growing and they believed in CSCS. We didn't have many applications that needed this level of scaling one year ago, but the situation is changing rapidly.

HPCwire: What's on the horizon for CSCS?

Sawley: We are the national center in a relatively small country, so we need to develop stronger awareness of CSCS and our scientific services. We expect from our large-scale system and services to make CSCS a more attractive partner for international collaborations. In a number of scientific fields, this country has outstanding researchers. You can't innovate without excellent researchers. I think the HPC Challenge is also a good fit when you want to innovate. You wouldn't use it in every situation and for every procurement, but it works well when you need to make a big leap forward. Having access to a leadership class scientific instrument such as an innovative high performance computer is a strong stimulus for new thinking.
 

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