GE Research Enters the Exascale Era

By Oliver Peckham

July 28, 2022

The pitch for GE Research is easy, as Richard Arthur, senior director of computational methods research for GE Research, explained at the latest meeting of the DOE’s Advanced Scientific Computing Advisory Committee (ASCAC): a third of the electrons in the world that flow through devices are generated on GE equipment; every two seconds, a plane takes off using GE engines; 16,000 medical scans are performed on GE equipment every minute — the list goes on. At the meeting, Arthur highlighted how crucial HPC resources — specifically, DOE-provided HPC resources — have been to the company’s gargantuan operations and research endeavors and highlighted how the exascale era stands to transform them even more.

GE’s equipment, Arthur said, is tightly regulated, has a long expected life and is very expensive. “When they’re not working, there are significant consequences,” he said. So, when GE Research runs models during the development of this equipment: “It is very important that these not be ‘garbage in, garbage out’ kinds of models.”

A lot of GE Research’s work occupies “the swim lane of computational fluid dynamics,” Arthur continued: the flow of gases through jet engines, the flow of water through hydroelectric plants, the flow of wind through wind turbines and so forth. In service of this work is a lot — a lot — of time and grants from the Department of Energy.

“By 2017,” Arthur said. “in peer-reviewed competitive grants, GE had already used over 1 billion core-hours on the National Lab machines across a number of the labs, and by today, that’s 14 INCITE awards, 22 ALCC [ASCR Leadership Computing Challenge] awards, 6 HPC4Manufacturing [awards], and other directors’ discretionary grants.”

Energy

The time, of course, is put to good use. Some years ago, GE — the largest supplier of gas turbines in the world — was struggling to isolate a problem affecting its turbines. “Before having access to these machines, we could only look at one combustor at a time,” Arthur said. “But there was an interaction between multiple combustors that was creating some problems that we would only be able to see once we were able to simulate multiple combustors.”

The necessary modification in GE’s gas turbines, courtesy of supercomputer simulations. Image courtesy of GE Research.

So GE received assistance from the National Labs, managing to simulate multiple combustors, isolate the problem (a fuel hole that needed to be moved by a fraction of an inch) and dramatically improve its turbine efficiency. (This work also won HPCwire’s 2016 Readers’ Choice Award for Best Use of HPC in Energy).

With regard to wind turbines, GE Research has been using a 2020 ALCC award to understand the impact of coastal low-level jets on wind energy. “We wanted to understand these effects so that we could predict what the performance of the wind turbine farm would be and understand the reliability of the operations,” Arthur said. Further, GE Research is now working to understand the wake effects that the first turbines to encounter wind have on the turbines further back — an effect that Arthur said could reduce efficiency by up to 40 percent, for which no physical wind tunnel is available for testing.

Image courtesy of GE Research.

“But [we’re] still not able to do complex interactions with things like wave swells and more complicated atmospheric phenomena [or] interactions between wind farms,” Arthur said. “With exascale, we believe that the wind farm-wind farm interactions become feasible [to simulate].”

Aviation

Another of the major efforts being undertaken by GE Research is the so-called “RISE” jet engine program, which stands for “Revolutionary Innovation for Sustainable Engines.” RISE is a joint venture between GE and French multinational Safran and the third in a line of aircraft modeling and improvement initiatives dating back to 1979. The previous generation of modeling, called LEAP, achieved a 15 percent increase in fuel efficiency for GE’s engine designs; RISE has a goal of a 20 to 100 percent reduction in carbon emissions and a further 20 percent reduction in fuel burn.

Image courtesy of GE Research.

“This is seen as a path to hydrogen [fuel],” Arthur said. “But to get the benefit of the zero carbon emissions out of hydrogen as a fuel, there are things you need to address.” Those include: price per kilometer (hydrogen fuel is expensive); increased volume per joule (hydrogen takes up more space); and, relatedly, decreased range (as less fuel can be stored). “In order to make hydrogen viable,” Arthur said, “we have to improve the propulsive efficiency significantly.” (RISE, Arthur later added, “is the path to hydrogen, but also to other things, such as open fan design as opposed to the nacelle.)

Doing this means conducting extremely detailed, unprecedented simulations of the engines in question. At the start of its simulation efforts many years ago, GE was only able to simulate one blade of a turbine at a time, which was of limited use. (“Really, what we care about in terms of the airflow is not the blade but the passage between two blades.”) That scaled up over time: first to multiple blades, then to multiple rows of blades, then to multiple stages of the engine. Simulating the full engine at full scale and exacting accuracy, Arthur explained, is the dream.

But, of course: “At any given time, we’ve only been able to simulate a problem as large as will fit on a given system,” Arthur said. Over the years, GE Research had leveraged a variety of systems for its turbine simulation work, ranging from Oak Ridge National Laboratory’s Jaguar system to Argonne National Laboratory’s Mira system to its own internal Cray systems. And, while these produced valuable results — Arthur said that the research on Mira drastically changed their approach to handling ice formation — the full engine simulation has remained elusive.

“Before we can run a flight test, we run a rig test,” Arthur said. “This is typically done in a wind tunnel, but these wind tunnels are smaller than the actual product. … We build this tiny version that will fit into NASA’s wind tunnel, and the relevant Reynolds numbers for the airflow of the small engine are significantly lower when compared to the fans that you see in the full product. So there is no wind tunnel that we can use to test the design at the product scale. So that is not tested until the flight test.”

So, when GE Research was dreaming of exascale back in 2019, some of their dreams included the use of exascale systems to reach another level of engine simulations.

“We laid some foundation for this with the 2021 INCITE [program] and this completed the first-of-their-kind wall-resolved large eddy simulations of the full 3D open fan at a high Reynolds number,” Arthur said. “This was still at rig test scale. … If we wanted to perform these same kinds of simulations at the flight scale, we would need a much bigger machine.”

“And this,” Arthur continued, “is where we approached Oak Ridge and ultimately put into the ALCC program this proposal to leverage basically everything we could possibly get access to.”

And the award was successful: significant time on the first official exascale system, Frontier, as well as time on Oak Ridge’s Summit system and NERSC’s Perlmutter system. (“And you can bet that if anybody leaves spare hours available we will be eagerly seeking to scoop those up,” Arthur said.)

“To run the takeoff test at the rig test scale, that simulation would be about 70,000 node hours on Summit,” Arthur added. “That same simulation at product scale … is over 6 million node hours on Summit.”

“So,” Arthur concluded, “we love Frontier!”

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