Skip to content

Codeplay inks landmark deal with U.S. government to enable next-generation supercomputer

The AI software firm will work with researchers in the high-performance compute community to analyze big data to simulate future pandemics and other areas.


Image: iStockphoto/niplot

The National Energy Research Scientific Computing Center at Lawrence Berkeley National Laboratory, in collaboration with the Argonne Leadership Computing Facility, is partnering with UK-based Codeplay Software to enhance GPU compiler capabilities for NVIDIA.

More about Innovation

This collaboration will help NERSC and ALCF users, along with researchers in the high-performance computing community, to produce high-performance applications that are portable across compute architectures from multiple vendors.

SEE: Research: Quantum computing in the enterprise; key vendors, anticipated benefits, and impact (TechRepublic Premium)

Today, most artificial intelligence software, including for cars, is developed using graphics processors designed for video games, according to Codeplay. The company provides tools designed to enable software to be accelerated by graphics processors or the latest specialized AI processors.

NVIDIA A100 GPUs will power NERSC’s next-generation supercomputer, Perlmutter, Codeplay said. NERSC supercomputers are used for scientific research by researchers working in diverse areas such as alternative energy, environment, high-energy and nuclear physics, advanced computing, materials science and chemistry.

Enabling treatments and strategies to combat the pandemic

Over the past year, 20 research teams have been involved in COVID-19 simulations for analysis and to develop solutions to combat the virus, Codeplay said. ALCF supercomputers enable scientific research and engineering by offering supercomputing resources and hands-on expertise to the research community.

These systems have helped advance science computing in an array of areas through the convergence of simulation, data science and machine learning methods.

The simulations have accelerated the development of treatments and strategies to help fight the COVID-19 virus.

SEE: The CIO’s guide to quantum computing (free PDF) (TechRepublic)

The power of open source

Today the SYCL open standard programming model supports a variety of accelerators through multiple implementations, Codeplay said. SYCL will be supported on the forthcoming Department of Energy Aurora exascale supercomputer and, with this work, it can be used with Perlmutter to help scientific app developers and users to be more productive, according to Codeplay.

“With thousands of users and a wide range of applications using NERSC’s resources, we must support a wide range of programming models,” explained Brandon Cook, application performance specialist at NERSC, in a statement. “In addition to directive-based approaches, we see modern C++ language-based approaches to accelerator programming, such as SYCL, as an important component of our programming environment offering for users of Perlmutter.”

“As a key programming model for Argonne’s upcoming exascale system, SYCL and DPC++ will benefit the broader DOE community by providing portability of accelerator programming models across DOE computing facilities,” said Andrew Richards, founder and CEO of Codeplay Software, in a statement.

“We are delighted to see the SYCL programming standard being embraced by the U.S. national labs and providing scientists developing accelerated C++ with a standardized software platform,” Richards said.

The use of high-performance computing (HPC) in data modeling, AI and analytics has already significantly exceeded expectations, Codeplay said. But the next decade will see explosive growth in capability and performance, achieved with special new processors and industry-standard software programmability. 

Today, many semiconductor and processor companies have their own specialized processor architecture tuned for complex AI functions. The latest AI applications are using neural networks to enable machine learning applications, and these processors enable NN operations to be performed with greater efficiency, according to the company.

However, as processor design tries to catch-up on AI needs, research continues to evolve, extending the processing needs of the underlying hardware and software can progress independently is crucial to compete.

Codeplay is “a big believer in open standards,” Richards said. The company has developed a range of products called ComputeSuite that aim to bridge the gap between the latest AI processors and AI application developers using established open standard interfaces.

Also see