Leading supercomputer sites choose Cerebras to accelerate AI

Leading supercomputer sites choose Cerebras to accelerate AI


Across the US and Europe, supercomputing centers rely on Cerebras chip-scale solutions for high-performance artificial intelligence.

Hamburg, Germany, May 31, 2022–(BUSINESS WIRE)–At ISC 2022, Cerebras Systems, a leader in high-performance artificial intelligence (AI) computing, shared news about several of its supercomputing partners including the European Parallel Computing Center (EPCC) and the Leibniz Supercomputing Center ( LRZ), Lawrence Livermore National Laboratory, Argonne National Laboratory (ANL), National Center for Supercomputing Applications (NCSA), and Pittsburgh Supercomputing Center (PSC).

“At Cerebras Systems, our goal is to revolutionize computing,” said Andrew Feldman, CEO and co-founder of Cerebras Systems. “It is exciting to see some of the most respected supercomputing centers around the world deploy CS-2 to accelerate AI workloads and achieve incredible scientific breakthroughs in climate research, precision medicine, computational fluid dynamics, and more.”

In Europe, EPCC in the UK and LRZ in Germany have announced the deployment of the CS-2 to accelerate scientific research in their regions. EPCC has been a customer since 2021, and recently approved the upgrade of the CS-1 to CS-2 after the success of the first system. CS-2 will significantly accelerate Natural Language Processing (NLP) projects and also expand its use in EPCC’s public health initiatives, making Edinburgh an AI hub for innovation in the region.

“EPCC has had a great experience with CS-1 and is happy to upgrade to CS-2. The support and engagement we have received from Cerebras has been fantastic, and we look forward to further success with our new system,” said Professor Mark Parsons, EPCC Director.

LRZ is set to accelerate scientific innovation and discovery in Germany with the CS-2 in its prospective AI supercomputer. The new supercomputer will go online this summer, enabling German researchers to boost scientific research and innovation using artificial intelligence. Initial work will focus on medical image processing that includes innovative algorithms for analyzing medical images, or computer-aided capabilities to speed diagnosis and diagnosis, and computational fluid dynamics (CFD) to advance understanding in areas such as aerospace engineering and manufacturing.

“Currently, we see that the demand for AI computing is doubling every three to four months with our users. With the high integration of processors, memory and networks integrated on a single chip, Cerebras enables high performance and speed. This promises significantly more efficiency in data processing and therefore faster penetration of results. scientific study,” said Professor Dr. Dieter Kranzlmueller, Director of the LRZ.

Continuing this trend, NCSA’s new HOLL-I supercomputer for large-scale machine learning is powered by CS-2. HOLL-I is unique in NCSA’s supercomputing portfolio in that it is designed to handle machine learning functions at intense speeds that will reduce computing time and lower overall costs, while obtaining exceptional performance results. Initial AI applications focus on the NCSA Group of Industry Partners; However, HOLL-I will be widely available for use in high-volume AI projects, at cost.

“We are excited to be running the Cerebras CS-2 system in the HOLL-I supercomputer,” said Vlad Kendratenko, Director of the Center for Artificial Intelligence Innovation at NCSA.

The Argonne National Laboratory’s Command Computing Facility was recently upgraded from one CS-1 system to two CS-2 systems. As part of the AI ​​Testbed, CS-2 Systems enables researchers to explore next-generation applications of machine learning and workloads to advance the use of AI in science. Previous work done on Cerebras CS-1 and CS-2 has been nominated for a special Gordon Bell Award for HPC-based COVID-19 research at SC21.

PSC has also doubled its AI capacity to 1.7 million AI cores, with two CS-2 systems, powering the Neocortex supercomputer at the center for high-performance AI. With CS-2 systems, Neocortex allows researchers to train larger deep learning models with larger data sets, while scaling parallelism to unprecedented levels.

“With our two Cerebras CS-2 systems, we look forward to the breakthroughs that will be enabled by the currently larger capabilities of Neocortex,” said Paula Buitrago, principal investigator of the neocortex and director of artificial intelligence and big data at PSC. “We will continue to work with the research community to help them take advantage of this technology that is more powerful by orders of magnitude.”

The Cerebras CS-2 is the fastest artificial intelligence system out there, powered by the largest processor ever built – the Cerebras Wafer-Scale Engine 2 (WSE-2), which is 56 times larger than the nearest competitor. As a result, CS-2 provides more AI-enhanced computing cores, faster memory, and greater fabric bandwidth than any other deep learning processor out there. It is designed to speed up deep learning workloads reducing response time by orders of magnitude.

With clients and partners in North America, Asia, Europe and the Middle East, Cerebras delivers leading-edge AI solutions to a growing list of customers in the enterprise, government, and high-performance computing sectors including GlaxoSmithKline, TotalEnergies, Leibniz Supercomputing Centre, National Supercomputing Application Center, nference, Laboratory Argonne National, Lawrence Livermore National Laboratory, Pittsburgh Supercomputing Center, Edinburgh Parallel Computing Center (EPCC), National Energy Technology Laboratory, and Tokyo Electronics.

For more information on the Cerebras CS-2 Scientific Computing System, please visit: https://cerebras.net/industry-scientific-computing/

About Cerebras . Systems

Cerebras Systems is a team of leading computer engineers, computer scientists, deep learning researchers, and engineers of all kinds. We’ve come together to build a new class of computer systems, designed with the unique purpose of accelerating AI and forever changing the future of AI work. Our flagship product, the CS-2 system powered by the world’s largest processor – 850,000 Cerebras WSE-2 cores, enables customers to accelerate deep learning work by orders of magnitude on top of GPUs.

View the source version on businesswire.com: https://www.businesswire.com/news/home/20220531005775/ar/


Media contact:
Kim Zimmer
[email protected]

#Leading #supercomputer #sites #choose #Cerebras #accelerate