Exascale Computing to Help Accelerate Drive for Clean Fusion Energy

By Jon Bashor, Lawrence Berkeley National Laboratory Computing Sciences

October 2, 2017

Editor’s note: One of the U.S. Exascale Computing Project’s mandates is to explain how exascale computing power will enhance scientific discovery and society broadly. This article from ECP not only examines the need for exascale computing power to advance research on fusion reactor design but it also highlights the potential for collaboration with industry partners who will require this kind of power.

For decades, scientists have struggled to create a clean, unlimited energy source here on Earth by recreating the conditions that drive our sun. Called a fusion reactor, the mechanism would use powerful magnetic fields to confine and compress gases four times as hot as our sun. By using the magnetic fields to squeeze the gases, the atoms would fuse and release more energy than was used to power the reactor. But to date, that has only worked in theory.

Achieving fusion energy production would benefit society by providing a power source that is non-polluting, renewable and using fuels such as the hydrogen isotopes found in seawater and boron isotopes found in minerals.

Early fusion research projects in the 1950s and ‘60s relied on building expensive magnetic devices, testing them and then building new ones and repeating the cycle. In the mid-1970s, fusion scientists began using powerful computers to simulate how the hot gases, called plasmas, would be heated, squeezed and fused to produce energy. It’s an extremely complex and difficult problem, one that some fusion researchers have likened to holding gelatin together with rubber bands.

Using supercomputers to model and simulate plasma behavior, scientists have made great strides toward building a working reactor. The next generation of supercomputers on the horizon, known as exascale systems, will bring the promise of fusion energy closer.

The best-known fusion reactor design is called a tokamak, in which a donut-shaped chamber is used to contain the hot gases, inside. Because the reactors are so expensive, only small-scale ones have been built. ITER, an international effort to build the largest-ever tokamak-in the south of France. The project, conceived in 1985, is now scheduled to have its first plasma experiments in 2025 and begin fusion experiments in 2035. The estimated cost is 14 billion euros, with the European Union and six other nations footing the bill.

Historically, fusion research around the world has been funded by governments due to the high cost and long-range nature of the work.

But in the Orange County foothills of Southern California, a private company is also pursuing fusion energy, but taking a far different path than that of ITER and other tokamaks. Tri Alpha Energy’s cylindrical reactor design is completely different in its design philosophy, geometry, fuels and method of heating the plasma, all built with a different funding model. Chief Science Officer Toshiki Tajima says their approach makes them mavericks in the fusion community.

But the one thing both ITER and similar projects and Tri Alpha Energy have consistently relied on is using high-performance computers to simulate conditions inside the reactor as they seek to overcome the challenges inherent in designing, building and operating a machine that will replicate the processes of the sun here on Earth.

As each generation of supercomputers has come online, fusion scientists have been able to study plasma conditions in greater detail, helping them understand how the plasma will behave, how it may lose energy and disrupt the reactions, and what can be done to create and maintain fusion. With exascale supercomputers that are 50 times more powerful than today’s top systems looming on the horizon, Tri Alpha Energy sees great possibilities in accelerating the development of their reactor design. Tajima is one of 18 members of the industry advisory council for the U.S. Department of Energy’s (DOE) Exascale Computing Project (ECP).

“We’re very excited by the promise of exascale computing – we are currently fund-raising for our next-generation machine, but we can build a simulated reactor using a very powerful computer, and for this we would certainly need exascale,” Tajima said. “This would help us accurately predict if our idea would work, and if it works as predicted, our investors would be encouraged to support construction of the real thing.”

The Tri Alpha Energy fusion model builds on the experience and expertise of Tajima and his longtime mentor, the late Norman Rostoker, a professor of physics at the University of California, Irvine (UCI). Tajima first met Rostoker as a graduate student, leaving Japan to study at Irvine in 1973. In addition to his work with TAE, Tajima holds the Norman Rostoker Chair in Applied Physics at UCI. In 1998, Rostoker co-founded TAE, which Tajima joined in 2011.

In it for the long run

It was also in the mid-1970s, that the U.S. Atomic Energy Commission, the forerunner of DOE, created a computing center to support magnetic fusion energy research, first with a cast-off computer from classified defense programs, but then with a series of ever-more capable supercomputers. From the outset, Tajima was an active user, and still remembers he was User No. 1100 at the Magnetic Fusion Energy Computer Center. The Control Data Corp. and Cray supercomputers were a big leap ahead of the IBM 360 he had been using.

“The behavior of plasma could not easily be predicted with computation back then and it was very hard to make any progress,” Tajima said. “I was one of the very early birds to foul up the machines. When the Cray-1 arrived, it was marvelous and I fell in love with it.”

At the time, the tokamak was seen as the hot design and most people in the field gravitated in this direction, Tajima said, and he followed. But after learning about plasma-driven accelerators under Professor Rostoker, in 1976 he went to UCLA to work with Prof. John Dawson. “He and I shared a vision of new accelerators and we began using large-scale computation in 1975, an area in which I wanted to learn more from him,” Tajima said.

As a result, the two men wrote a paper entitled “Laser Electron Accelerator,” which appeared in Physical Review Letters in 1979. The seminal paper explained how firing an intense electromagnetic pulse (or beam of particles) into a plasma can create a wake in the plasma and that electrons, and perhaps ions, trapped in this wake can be accelerated to very high energies.

TAE’s philosophy, built on Rostoker’s ideas, is to combine both accelerator and fusion plasma research. In a tokamak, the deuterium-tritium fuel needs to be heated and confined at an energy level of 10,000 eV (electron volts) for fusion to occur. The TAE reactor, however, needs to be 30 times hotter. In a tokamak, the same magnetic fields that confine the plasma also heat it to 3 billion degrees C. In the TAE machine, the energy will be injected using a particle accelerator. “A 100,000 eV beam is nothing for an accelerator,” Tajima said, pointing to the 1G eV BELLA device at DOE’s Lawrence Berkeley National Laboratory. “Using a beam-driven plasma is relatively easy but it may be counterintuitive that you can get higher energy with more stability — the more energetic the wake is, the more stable it becomes.”

But this approach is not without risk. With the tokamak, the magnetic fields protect the plasma, much like the exoskeleton of a beetle protects the insect’s innards, Tajima said. But the accelerator beam creates a kind of spine, which creates the plasma by its weak magnetic fields, a condition known as Reverse Field Configuration. One of Rostoker’s concerns was that the plasma would be too vulnerable to other forces in the early stages of its formation. However, in the 40-centimeter diameter cylindrical reactor, the beam forms a ring like a bicycle tire, and like a bicycle, the stability increases the faster the wheels spin.

“The stronger the beam is, the more stable the plasma becomes,” Tajima said. “This was the riskiest problem for us to solve, but in early 2000 we showed the plasma could survive and this reassured our investors. We call this approach of tackling the hardest problem first ‘fail fast’.”

Another advantage of TAE’s approach is that the main fuel, Boron-11, does not produce neutrons as a by-product; instead it produces three alpha particles, which is the basis of the company’s name. A tokamak, using hydrogen-isotope fuels, generates neutrons, which can penetrate and damage materials, including the superconducting magnets that confine the tokamak plasma. To prevent this, the tokamak reactor requires one-meter-thick shielding. Without the need to contain neutrons, the TAE reactor does not need heavy shielding. This also helps reduce construction costs.

Computation Critical to Future Progress

With his 40 years of experience using HPC to advance fusion energy, Tajima offers a long-term perspective, from the past decades to exascale systems in the early 2020s. As a principal investigator on the Numerical Tokamak project in the early 1990s, he has helped build much of the HPC ecosystem for fusion research.

At the early stage of modeling fusion behavior, the codes focus on the global plasma at very fast time scales. These codes, known as MHD codes (magnetohydrodynamics), are not as computationally “expensive,” meaning they do not require as many computing resources, and at TAE were run on in-house clusters.

The next step is to model the more minute part of the plasma instability, known as kinetic instability, which requires more sophisticated codes that can simulate the plasma in greater detail over longer time scales. Achieving this requires more sophisticated systems. Around 2008-09, TAE stabilized this stage of the problem using its own computing system and by working with university collaborators who have access to federally funded supercomputing centers, such as those supported by DOE. “Our computing became more demanding during this time,” Tajima said.

The third step, which TAE is now tackling, is to make a plasma that can “live” longer, which is known as the transport issue in the fusion community. “This is a very, very difficult problem and consumes large amounts of computing resources as it encompasses a different element of the plasma,” Tajima said, “and the plasma becomes much more complex.”

The problem involves three distinct functions:

  • The core of the field reverse configuration, which is where the plasma is at the highest temperature
  • The “scrape-off layer,” which is the protective outer layer of ash on the core and which Tajima likens to an onion’s skin
  • The “ash cans,” or diverters, that are at each end of the reactor. They remove the ash, or impurities, from the scrape-off layer, which can make the plasma muddy and cause it to behave improperly.

“The problem is that the three elements behave very, very differently in both the plasma physics as well as in other properties,” Tajima said. “For example, the diverters are facing the metallic walls so you have to understand the interaction of the cold plate metals and the out-rushing impurities. And those dynamics are totally different than the core which is very high temperature and very high energy and spinning around like a bicycle tire, and the scrape-off layer.”

These factors are all coupled to each other using very complex geometries and in order to see if the TAE approach is feasible, researchers need to simulate the entirety of the reactor in order to understand and eventually control the reactions.

“We will run a three-layered simulation of our fusion reactor on the computer, with the huge particle code, the transport code and the neural net on the simulation – that’s our vision and we will certainly need an exascale machine to do this,” Tajima said. “This will allow us to predict if our concept works or not in advance of building machine so that our investors’ funds are not wasted.”

The overall code will have three components. At the basic level will be a representative simulation of particles in each part of the plasma. The second layer will be the more abstract transport code, which tracks heat moving in and out of the plasma. But even on exascale systems, the transport code will not be able to run fast enough to keep up with real-time changes in the plasma. Instabilities which affect the heat transport in the plasma come and go in milliseconds.

“So, we need a third layer that will be an artificial neural net, which will be able to react in microseconds, which is a bit similar to a driverless auto, and will ‘learn’ how to control the bicycle tire-shaped plasma, Tajima said. This application will be run on top of transport code and it will observe experimental data and react appropriately to keep the simulation running.

“Doing this will certainly require exascale computing,” Tajima said. “Without it we will take up to 30 years to finish – and our investors cannot wait that long. This project has been independent of the government funding, so that our investors’ fund provided an independent, totally different path toward fusion. This could amount to a means of national security to provide an alternative solution to a problem as large as fusion energy. Society will also benefit from a clean source of energy and our exascale-driven reactor march will be a very good thing for the nation and the world.”

Advanced Accelerators are Pivotal

Both particle accelerators and fusion energy are technologies important to the nation’s scientific leadership, with research funded over many decades by the Department of Energy and its predecessor agencies.

Not only are particle accelerators a vital part of the DOE-supported infrastructure of discovery science and university research, they also have private-sector applications and a broad range of benefits to industry, security, energy, the environment and medicine.

Since Toshiki Tajima and John Dawson published their paper “Laser Electron Accelerator” in 1979, the idea of building smaller accelerators, with the length measure in meters instead of kilometers, has gained traction. In these new accelerators, particles “surf” in the plasma wake of injected particles, reaching very high energy levels in very short distances.

According to Jean-Luc Vay, a researcher at DOE’s Lawrence Berkeley National Laboratory, taking full advantage of accelerators’ societal benefits, game-changing improvements in the size and cost of accelerators are needed. Plasma-based particle accelerators stand apart in their potential for these improvements, according to Vay, and turning this from a promising technology into a mainstream scientific tool depends critically on high-performance, high-fidelity modeling of complex processes that develop over a wide range of space and time scales.

To help achieve this goal, Vay is leading a project called Exascale Modeling of Advanced Particle Accelerators as part of DOE’s Exascale Computing Project. This project supports the practical economic design of smaller, less-expensive plasma-based accelerators.

As Tri Alpha Energy pursues its goal of using a particle accelerator (though this accelerator is not related to wakefield accelerators) to achieve fusion energy, the company is also planning to apply its experience and expertise in accelerator research for medical applications. Not only will this effort produce returns for the company’s investors, but it should also help advance TAE’s understanding of accelerators and using them to create a fusion reactor.

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!

The Case for an Edge-Driven Future for Supercomputing

September 24, 2021

“Exascale only becomes valuable when it’s creating and using data that we care about,” said Pete Beckman, co-director of the Northwestern-Argonne Institute of Science and Engineering (NAISE), at the most recent HPC Read more…

Three Universities Team for NSF-Funded ‘ACES’ Reconfigurable Supercomputer Prototype

September 23, 2021

As Moore’s law slows, HPC developers are increasingly looking for speed gains in specialized code and specialized hardware – but this specialization, in turn, can make testing and deploying code trickier than ever. Now, researchers from Texas A&M University, the University of Illinois at Urbana... Read more…

Qubit Stream: Monte Carlo Advance, Infosys Joins the Fray, D-Wave Meeting Plans, and More

September 23, 2021

It seems the stream of quantum computing reports never ceases. This week – IonQ and Goldman Sachs tackle Monte Carlo on quantum hardware, Cambridge Quantum pushes chemistry calculations forward, D-Wave prepares for its Read more…

Asetek Announces It Is Exiting HPC to Protect Future Profitability

September 22, 2021

Liquid cooling specialist Asetek, well-known in HPC circles for its direct-to-chip cooling technology that is inside some of the fastest supercomputers in the world, announced today that it is exiting the HPC space amid multiple supply chain issues related to the pandemic. Although pandemic supply chain... Read more…

TACC Supercomputer Delves Into Protein Interactions

September 22, 2021

Adenosine triphosphate (ATP) is a compound used to funnel energy from mitochondria to other parts of the cell, enabling energy-driven functions like muscle contractions. For ATP to flow, though, the interaction between the hexokinase-II (HKII) enzyme and the proteins found in a specific channel on the mitochondria’s outer membrane. Now, simulations conducted on supercomputers at the Texas Advanced Computing Center (TACC) have simulated... Read more…

AWS Solution Channel

Introducing AWS ParallelCluster 3

Running HPC workloads, like computational fluid dynamics (CFD), molecular dynamics, or weather forecasting typically involves a lot of moving parts. You need a hundreds or thousands of compute cores, a job scheduler for keeping them fed, a shared file system that’s tuned for throughput or IOPS (or both), loads of libraries, a fast network, and a head node to make sense of all this. Read more…

The Latest MLPerf Inference Results: Nvidia GPUs Hold Sway but Here Come CPUs and Intel

September 22, 2021

The latest round of MLPerf inference benchmark (v 1.1) results was released today and Nvidia again dominated, sweeping the top spots in the closed (apples-to-apples) datacenter and edge categories. Perhaps more interesti Read more…

The Case for an Edge-Driven Future for Supercomputing

September 24, 2021

“Exascale only becomes valuable when it’s creating and using data that we care about,” said Pete Beckman, co-director of the Northwestern-Argonne Institut Read more…

Three Universities Team for NSF-Funded ‘ACES’ Reconfigurable Supercomputer Prototype

September 23, 2021

As Moore’s law slows, HPC developers are increasingly looking for speed gains in specialized code and specialized hardware – but this specialization, in turn, can make testing and deploying code trickier than ever. Now, researchers from Texas A&M University, the University of Illinois at Urbana... Read more…

Qubit Stream: Monte Carlo Advance, Infosys Joins the Fray, D-Wave Meeting Plans, and More

September 23, 2021

It seems the stream of quantum computing reports never ceases. This week – IonQ and Goldman Sachs tackle Monte Carlo on quantum hardware, Cambridge Quantum pu Read more…

Asetek Announces It Is Exiting HPC to Protect Future Profitability

September 22, 2021

Liquid cooling specialist Asetek, well-known in HPC circles for its direct-to-chip cooling technology that is inside some of the fastest supercomputers in the world, announced today that it is exiting the HPC space amid multiple supply chain issues related to the pandemic. Although pandemic supply chain... Read more…

TACC Supercomputer Delves Into Protein Interactions

September 22, 2021

Adenosine triphosphate (ATP) is a compound used to funnel energy from mitochondria to other parts of the cell, enabling energy-driven functions like muscle contractions. For ATP to flow, though, the interaction between the hexokinase-II (HKII) enzyme and the proteins found in a specific channel on the mitochondria’s outer membrane. Now, simulations conducted on supercomputers at the Texas Advanced Computing Center (TACC) have simulated... Read more…

The Latest MLPerf Inference Results: Nvidia GPUs Hold Sway but Here Come CPUs and Intel

September 22, 2021

The latest round of MLPerf inference benchmark (v 1.1) results was released today and Nvidia again dominated, sweeping the top spots in the closed (apples-to-ap Read more…

Why HPC Storage Matters More Now Than Ever: Analyst Q&A

September 17, 2021

With soaring data volumes and insatiable computing driving nearly every facet of economic, social and scientific progress, data storage is seizing the spotlight. Hyperion Research analyst and noted storage expert Mark Nossokoff looks at key storage trends in the context of the evolving HPC (and AI) landscape... Read more…

GigaIO Gets $14.7M in Series B Funding to Expand Its Composable Fabric Technology to Customers

September 16, 2021

Just before the COVID-19 pandemic began in March 2020, GigaIO introduced its Universal Composable Fabric technology, which allows enterprises to bring together Read more…

Ahead of ‘Dojo,’ Tesla Reveals Its Massive Precursor Supercomputer

June 22, 2021

In spring 2019, Tesla made cryptic reference to a project called Dojo, a “super-powerful training computer” for video data processing. Then, in summer 2020, Tesla CEO Elon Musk tweeted: “Tesla is developing a [neural network] training computer called Dojo to process truly vast amounts of video data. It’s a beast! … A truly useful exaflop at de facto FP32.” Read more…

Enter Dojo: Tesla Reveals Design for Modular Supercomputer & D1 Chip

August 20, 2021

Two months ago, Tesla revealed a massive GPU cluster that it said was “roughly the number five supercomputer in the world,” and which was just a precursor to Tesla’s real supercomputing moonshot: the long-rumored, little-detailed Dojo system. “We’ve been scaling our neural network training compute dramatically over the last few years,” said Milan Kovac, Tesla’s director of autopilot engineering. Read more…

Esperanto, Silicon in Hand, Champions the Efficiency of Its 1,092-Core RISC-V Chip

August 27, 2021

Esperanto Technologies made waves last December when it announced ET-SoC-1, a new RISC-V-based chip aimed at machine learning that packed nearly 1,100 cores onto a package small enough to fit six times over on a single PCIe card. Now, Esperanto is back, silicon in-hand and taking aim... Read more…

CentOS Replacement Rocky Linux Is Now in GA and Under Independent Control

June 21, 2021

The Rocky Enterprise Software Foundation (RESF) is announcing the general availability of Rocky Linux, release 8.4, designed as a drop-in replacement for the soon-to-be discontinued CentOS. The GA release is launching six-and-a-half months after Red Hat deprecated its support for the widely popular, free CentOS server operating system. The Rocky Linux development effort... Read more…

Berkeley Lab Debuts Perlmutter, World’s Fastest AI Supercomputer

May 27, 2021

A ribbon-cutting ceremony held virtually at Berkeley Lab's National Energy Research Scientific Computing Center (NERSC) today marked the official launch of Perlmutter – aka NERSC-9 – the GPU-accelerated supercomputer built by HPE in partnership with Nvidia and AMD. Read more…

Intel Completes LLVM Adoption; Will End Updates to Classic C/C++ Compilers in Future

August 10, 2021

Intel reported in a blog this week that its adoption of the open source LLVM architecture for Intel’s C/C++ compiler is complete. The transition is part of In 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…

Hot Chips: Here Come the DPUs and IPUs from Arm, Nvidia and Intel

August 25, 2021

The emergence of data processing units (DPU) and infrastructure processing units (IPU) as potentially important pieces in cloud and datacenter architectures was Read more…

Leading Solution Providers

Contributors

AMD-Xilinx Deal Gains UK, EU Approvals — China’s Decision Still Pending

July 1, 2021

AMD’s planned acquisition of FPGA maker Xilinx is now in the hands of Chinese regulators after needed antitrust approvals for the $35 billion deal were receiv Read more…

10nm, 7nm, 5nm…. Should the Chip Nanometer Metric Be Replaced?

June 1, 2020

The biggest cool factor in server chips is the nanometer. AMD beating Intel to a CPU built on a 7nm process node* – with 5nm and 3nm on the way – has been i Read more…

HPE Wins $2B GreenLake HPC-as-a-Service Deal with NSA

September 1, 2021

In the heated, oft-contentious, government IT space, HPE has won a massive $2 billion contract to provide HPC and AI services to the United States’ National Security Agency (NSA). Following on the heels of the now-canceled $10 billion JEDI contract (reissued as JWCC) and a $10 billion... 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…

Quantum Roundup: IBM, Rigetti, Phasecraft, Oxford QC, China, and More

July 13, 2021

IBM yesterday announced a proof for a quantum ML algorithm. A week ago, it unveiled a new topology for its quantum processors. Last Friday, the Technical Univer Read more…

Intel Launches 10nm ‘Ice Lake’ Datacenter CPU with Up to 40 Cores

April 6, 2021

The wait is over. Today Intel officially launched its 10nm datacenter CPU, the third-generation Intel Xeon Scalable processor, codenamed Ice Lake. With up to 40 Read more…

Frontier to Meet 20MW Exascale Power Target Set by DARPA in 2008

July 14, 2021

After more than a decade of planning, the United States’ first exascale computer, Frontier, is set to arrive at Oak Ridge National Laboratory (ORNL) later this year. Crossing this “1,000x” horizon required overcoming four major challenges: power demand, reliability, extreme parallelism and data movement. Read more…

Intel Unveils New Node Names; Sapphire Rapids Is Now an ‘Intel 7’ CPU

July 27, 2021

What's a preeminent chip company to do when its process node technology lags the competition by (roughly) one generation, but outmoded naming conventions make it seem like it's two nodes behind? For Intel, the response was to change how it refers to its nodes with the aim of better reflecting its positioning within the leadership semiconductor manufacturing space. Intel revealed its new node nomenclature, and... Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire