Atos Invokes Sovereignty, Sustainability at Next-Gen Supercomputer Launch

By Oliver Peckham

February 17, 2022

At an elaborate event yesterday, French computing firm Atos announced its newest supercomputer: the BullSequana XH3000, a system that the company can scale to an exaflops performance and which works with hardware from AMD, Intel, Nvidia and (when it arrives) SiPearl. (More feature coverage on the technical aspects of the XH3000 is available here.) At the event – which included a series of keynotes and two roundtables – Atos and its associates hammered home two key ideas that were repeated across nearly every speaker.

Those themes: sovereignty and sustainability.

Sovereignty

The European Union has been working to insource more component design and production for some years now, including major efforts like the European Processor Initiative (EPI) and, most recently, the European Chips Act, which – if passed – would direct nearly $50 billion in funding toward building up the European semiconductor industry through 2030. Supply chain disruptions (which factored into the design of the Chips Act) have only amplified those desires for a more sovereign European computing ecosystem, with Atos and SiPearl standing as the corporate representatives of that envisioned future. At the event today, that vision was fortified: Europe wants computing sovereignty, and Atos is the steward of the throne.

One of the opening talks was from Cédric O, state secretary of digital transition and electronic communication for the French government. “There is a strong urgency as far as Europe is concerned to get back into the technological race,” he said. “We do believe that we have to be able to build a global landscape, but Europe has to be able to carve its own path,” he added, citing growing competition from the United States and China. Atos and its partners, he said, were “sovereign assets.”

Rodolphe Belmer, CEO of Atos, was more than happy to agree with this notion. “As the undisputed European leader in supercomputing, Atos plays a key role in addressing sovereignty issues, ensuring that key HPC knowledge and skills are fostered and remain on our continent,” he said, calling HPC a “pillar of the digital autonomy and digital superiority of states.” He touted Atos’ successes in the Top500 (which includes 40 systems from Atos) and in recent EuroHPC projects like Deucalion (Portugal), Discoverer (Bulgaria), MeluXina (Luxembourg) and the forthcoming Leonardo pre-exascale system (Italy).

Left to right: Eric Eppe, Sanzio Bassini, Jean-Philippe Nominé, Thomas Lippert, and panel moderator Addison Snell, CEO of Intersect360 Research.

Atos’ position was elaborated upon by Arnaud Bertrand (fellow, senior vice president and head of big data and security strategy, innovation and R&D for Atos) during one of the roundtables. “This is where our technological business meets geopolitics and public policy, and what makes us truly enthusiastic to work on these topics at the crossroads of those diverse domains,” he said. He added that geopolitical instability created challenges for the European computing sector, and those sectors that rely upon it. “The fact that Taiwan is facing some risk is making some risk for our industry,” he said, by way of example. “America is already addressing the issue, and Europe has the will and the means to do it.” Specifically, he cited the European Chip Act as “key for the future of our industry.”

“The pandemic reminded everyone that it was important to be sovereign,” added Eric Eppe, head of portfolio and solutions for HPC and quantum at Atos. “And everyone wants to be sovereign to control [their] own destiny. It’s not about not being able to use U.S.-based types of CPUs or GPUs—it’s not that. … Really, you want to be able to do your own studies, discover new drugs, model the climate changes … We want to be sovereign in doing that. It’s about freedom, it’s not about technology.”


“The pandemic reminded everyone that it was important to be sovereign.”


But Atos wasn’t the only one advocating for that sense of freedom—far from it. Several of the other panelists echoed the need for European computing sovereignty, though some sought to further clarify what, exactly, the word meant.

Jean-Philippe Nominé – a fellow and HPC strategic collaborations manager for French research organization CEA – said that it wasn’t about “running 100 percent indigenous technology” (“It would not be possible, anyway”). Instead, he said, “What is important is that you are not prevented by others to do what you want to do — that you keep the control of what you are doing. And for this you must master a number of critical skills and technologies.”

Nominé broke those down into four levels, plus a bonus: first, he said, was mastering system integration and delivery – something that Atos had been doing for 20 years. Second was deeper work with the subsystems and components, such as participating in codesign of the interconnect of a system. Third was processor design, such as that being done by EPI. Fourth, he said, was the foundry level (“When you take a look at the EU Chips Act, there might be opportunities if this works in the coming years”). The bonus level: “sovereignty over applications,” Nominé said, leaning into his microphone, “does matter as well.”

“The processor—that’s really fundamental,” added Thomas Lippert, head of the Jülich Supercomputing Centre (FZJ). “If we don’t understand that, we never will have what we call in German ‘Augenhöhe’ – you are on the same level with your eyes as the other one. This I would call sovereignty, so that we can act as the others also can act and we can be competitive. That’s the important point.”

“The point is not for dominance,” agreed Sanzio Bassini, director of supercomputing applications and innovation for CINECA, Italy’s supercomputing organization.

Sustainability

For many of the participants – and, particularly, for Atos – this drive for sovereignty was complemented by a strongly perceived need for sustainability, which pervaded the event and was the focus of one of the roundtables. “[Atos is] looking at the carbon footprint at all the stages,” Bertrand said. “First at the design, then at the manufacture, third at the run phase and then at the dismantling phase.”

“At the design phase,” he continued, “we are looking component-by-component at the carbon footprint of each and every component composing this piece of technology. … Then, at the manufacturing [phase], we are looking at using [as little] carbon as possible, [focusing] on the transportation because the main cost of a supercomputer [outside] of the run phase is due to transportation of components and transportation of the system. So we are using, as much as we can, local manufacturing – both for sovereignty issues but also for [the] low carbon footprint impact.”

“Then,” he said, “we manufacture with a full process managing the carbon footprint of the manufacturing of the system. And then we use transportation, again, with a low carbon footprint. And then, at the run phase, we – through software – start to profile the applications to know [the] impact of each and every job. And our goal here is first to know the carbon footprint of a job in an HPC system, and then to be able to act—so to, potentially, re-parameter the system, change the way the workloads are run, to optimize the workloads, either for performance or for lower carbon footprint. And then we also handle the dismantling phase—which is, by the way, very important for the whole carbon footprint.”

The goal, Bertrand said: “100 percent carbon neutrality.”

“Of course, we won’t create an HPC [system] which is not consuming electricity and just powered by the wind—that’s not possible today (not yet),” he added. “It will be a huge challenge for this 100 percent carbon neutrality.” But: “We are working on it.”


“It will be a huge challenge for this 100 percent carbon neutrality.”


To that end, Belmer had pointed out earlier that all 40 of Atos’ machines in the Top500 were also ranked on the Green500 list, which ranks submitted supercomputers by flops per watt. Atos also revealed that the XH3000—which it expects to ship around Q4—provides “six times more performance per square meter than the previous generation” and hinted that the XH3000 was anticipated to have a longer lifespan (around six years).

Steve Conway, cofounder of Hyperion Research, pointed out the practical advantages of pursuing these sorts of efficiency improvements, noting that a decade ago, exascale computing was expected to require around 120MW of power; now, he said, the expectation is around 20MW. “That’s really, really important, because a 120MW exascale supercomputer would in some ways be very impractical for governments to fund,” he said. “This means the game can continue. This means computing progress can continue.”

Left to right: Arnaud Bertrand, Steve Conway, Alison Kennedy and panel moderator Ineke Vermeulen, senior marketing manager for Atos.

Alison Kennedy, director of the Science and Technology Facilities Council’s Hartree Center in the UK, shared another advantage of pursuing sustainability, saying that she had been seeing environmental concerns influence how younger candidates sought jobs. “We have to be conscious of the image that supercomputing has and to portray the importance of net zero and the steps that we are taking to support that,” she said.

For Nominé, the question was less about minimization of resource use and more about optimization.

“Let’s be clear,” he said. “I think HPC is not necessarily … here to be frugal, strictly speaking. We will probably go on using megawatts, or a few tens of megawatts – we will reach a plateau, by the way … we will probably not be able to grow forever. So let’s say we’re at this 10, 20 megawatts – the point is to make the best possible efficient use of those resources. To the maximum density of data processing or compute cycles to serve applications within this budget.”

Nominé cited the use of waste heat to heat the towns around HPC datacenters. “We have the watts, we have the flops … We want to use them!” he said. “We want to burn the megawatts! But we want to make the best possible use of that, and the remaining heat, if we can … reuse it, it’s nice. And I think there’s a lot of room in-between.”

Synergy

These twin goals – sovereignty and sustainability – were not portrayed as oppositional to each other: quite the opposite. “Supporting Atos and being able to accelerate as far as the technology is concerned and the footprint of Atos is concerned is … a sovereign, but also … an environmental question and challenge for Europe,” O said.

“We fight hard to build technological superiority,” Belmer agreed. “But this should not be at the detriment of our global warming engagements. This is why our supercomputers and edge servers are conceived, also, to minimize energy consumption and lower carbon emissions.”

But it was Bertrand, perhaps, who best drew the thematic throughline between sovereignty and sustainability, invoking Back to the Future’s Doc Brown while laying out a vision of the coming decades in European supercomputing.

“‘Your future is whatever you make it,’” he quoted. “‘So make it a good one!’”

To learn more about the XH3000, read the additional feature coverage from HPCwire here.

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!

The New MLPerf Storage Benchmark Runs Without ML Accelerators

October 3, 2024

MLCommons is known for its independent Machine Learning (ML) benchmarks.  These benchmarks have focused on mathematical ML operations and accelerators (e.g., Nvidia GPUs). Recently, MLCommons introduced the results of i Read more…

DataPelago Unveils Universal Engine to Unite Big Data, Advanced Analytics, HPC, and AI Workloads

October 3, 2024

DataPelago today emerged from stealth with a new virtualization layer that it says will allow users to move AI, data analytics, and ETL workloads to whatever physical processor they want, without making code changes, the Read more…

IBM Quantum Summit Evolves into Developer Conference

October 2, 2024

Instead of its usual quantum summit this year, IBM will hold its first IBM Quantum Developer Conference which the company is calling, “an exclusive, first-of-its-kind.” It’s planned as an in-person conference at th Read more…

Stayin’ Alive: Intel’s Falcon Shores GPU Will Survive Restructuring

October 2, 2024

Intel's upcoming Falcon Shores GPU will survive the brutal cost-cutting measures as part of its "next phase of transformation." An Intel spokeswoman confirmed that the company will release Falcon Shores as a GPU. The com Read more…

Texas A&M HPRC at PEARC24: Building the National CI Workforce

October 1, 2024

Texas A&M High-Performance Research Computing (HPRC) significantly contributed to the PEARC24 (Practice & Experience in Advanced Research Computing 2024) conference. Eleven HPRC and ACES’ (Accelerating Computin Read more…

A Q&A with Quantum Systems Accelerator Director Bert de Jong

September 30, 2024

Quantum technologies may still be in development, but these systems are evolving rapidly and existing prototypes are already making a big impact on science and industry. One of the major hubs of quantum R&D is the Q Read more…

The New MLPerf Storage Benchmark Runs Without ML Accelerators

October 3, 2024

MLCommons is known for its independent Machine Learning (ML) benchmarks.  These benchmarks have focused on mathematical ML operations and accelerators (e.g., N Read more…

DataPelago Unveils Universal Engine to Unite Big Data, Advanced Analytics, HPC, and AI Workloads

October 3, 2024

DataPelago today emerged from stealth with a new virtualization layer that it says will allow users to move AI, data analytics, and ETL workloads to whatever ph Read more…

Stayin’ Alive: Intel’s Falcon Shores GPU Will Survive Restructuring

October 2, 2024

Intel's upcoming Falcon Shores GPU will survive the brutal cost-cutting measures as part of its "next phase of transformation." An Intel spokeswoman confirmed t Read more…

How GenAI Will Impact Jobs In the Real World

September 30, 2024

There’s been a lot of fear, uncertainty, and doubt (FUD) about the potential for generative AI to take people’s jobs. The capability of large language model Read more…

IBM and NASA Launch Open-Source AI Model for Advanced Climate and Weather Research

September 25, 2024

IBM and NASA have developed a new AI foundation model for a wide range of climate and weather applications, with contributions from the Department of Energy’s Read more…

Intel Customizing Granite Rapids Server Chips for Nvidia GPUs

September 25, 2024

Intel is now customizing its latest Xeon 6 server chips for use with Nvidia's GPUs that dominate the AI landscape. The chipmaker's new Xeon 6 chips, also called Read more…

Building the Quantum Economy — Chicago Style

September 24, 2024

Will there be regional winner in the global quantum economy sweepstakes? With visions of Silicon Valley’s iconic success in electronics and Boston/Cambridge� Read more…

How GPUs Are Embedded in the HPC Landscape

September 23, 2024

Grasping the basics of Graphics Processing Unit (GPU) architecture is crucial for understanding how these powerful processors function, particularly in high-per Read more…

Shutterstock_2176157037

Intel’s Falcon Shores Future Looks Bleak as It Concedes AI Training to GPU Rivals

September 17, 2024

Intel's Falcon Shores future looks bleak as it concedes AI training to GPU rivals On Monday, Intel sent a letter to employees detailing its comeback plan after Read more…

Nvidia Shipped 3.76 Million Data-center GPUs in 2023, According to Study

June 10, 2024

Nvidia had an explosive 2023 in data-center GPU shipments, which totaled roughly 3.76 million units, according to a study conducted by semiconductor analyst fir Read more…

AMD Clears Up Messy GPU Roadmap, Upgrades Chips Annually

June 3, 2024

In the world of AI, there's a desperate search for an alternative to Nvidia's GPUs, and AMD is stepping up to the plate. AMD detailed its updated GPU roadmap, w Read more…

Granite Rapids HPC Benchmarks: I’m Thinking Intel Is Back (Updated)

September 25, 2024

Waiting is the hardest part. In the fall of 2023, HPCwire wrote about the new diverging Xeon processor strategy from Intel. Instead of a on-size-fits all approa Read more…

Ansys Fluent® Adds AMD Instinct™ MI200 and MI300 Acceleration to Power CFD Simulations

September 23, 2024

Ansys Fluent® is well-known in the commercial computational fluid dynamics (CFD) space and is praised for its versatility as a general-purpose solver. Its impr Read more…

Shutterstock_1687123447

Nvidia Economics: Make $5-$7 for Every $1 Spent on GPUs

June 30, 2024

Nvidia is saying that companies could make $5 to $7 for every $1 invested in GPUs over a four-year period. Customers are investing billions in new Nvidia hardwa Read more…

Shutterstock 1024337068

Researchers Benchmark Nvidia’s GH200 Supercomputing Chips

September 4, 2024

Nvidia is putting its GH200 chips in European supercomputers, and researchers are getting their hands on those systems and releasing research papers with perfor Read more…

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…

Leading Solution Providers

Contributors

Everyone Except Nvidia Forms Ultra Accelerator Link (UALink) Consortium

May 30, 2024

Consider the GPU. An island of SIMD greatness that makes light work of matrix math. Originally designed to rapidly paint dots on a computer monitor, it was then Read more…

Quantum and AI: Navigating the Resource Challenge

September 18, 2024

Rapid advancements in quantum computing are bringing a new era of technological possibilities. However, as quantum technology progresses, there are growing conc Read more…

IBM Develops New Quantum Benchmarking Tool — Benchpress

September 26, 2024

Benchmarking is an important topic in quantum computing. There’s consensus it’s needed but opinions vary widely on how to go about it. Last week, IBM introd Read more…

Google’s DataGemma Tackles AI Hallucination

September 18, 2024

The rapid evolution of large language models (LLMs) has fueled significant advancement in AI, enabling these systems to analyze text, generate summaries, sugges Read more…

Microsoft, Quantinuum Use Hybrid Workflow to Simulate Catalyst

September 13, 2024

Microsoft and Quantinuum reported the ability to create 12 logical qubits on Quantinuum's H2 trapped ion system this week and also reported using two logical qu Read more…

IonQ Plots Path to Commercial (Quantum) Advantage

July 2, 2024

IonQ, the trapped ion quantum computing specialist, delivered a progress report last week firming up 2024/25 product goals and reviewing its technology roadmap. Read more…

Intel Customizing Granite Rapids Server Chips for Nvidia GPUs

September 25, 2024

Intel is now customizing its latest Xeon 6 server chips for use with Nvidia's GPUs that dominate the AI landscape. The chipmaker's new Xeon 6 chips, also called Read more…

US Implements Controls on Quantum Computing and other Technologies

September 27, 2024

Yesterday the Commerce Department announced  export controls on quantum computing technologies as well as new controls for advanced semiconductors and additiv Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire