It is the largest AI processor ever made.

Cerebras Systems, which is building a new class of computers to accelerate artificial intelligence work by orders of magnitude beyond what is currently possible, announced Tuesday the launch of the Wafer Scale Engine 2 (WSE-2), the largest processor ever made. The new chip will power the Cerebras CS-2, the industry’s fastest AI computer, which will now run at 2X the speed of the first-generation CS-1. (BusinessWire)

WSE-2

The WSE-2 chip is tailor-made for AI processing. Manufactured by TSMC (TPE: 2330) on its 7nm-node, it is a single chip that is really an entire wafer boasting 2.6 trillion transistors and 850,000 AI optimized cores.

Compared to the first-generation version, the WSE-2 chip more than doubles all performance parameters on the chip – the transistor count, core count, memory, memory bandwidth, and fabric bandwidth.

In comparison, the largest graphics processor unit (GPU) has only 54 billion transistors – 2.55 trillion fewer transistors than the WSE-2. The WSE-2 also has 123x more cores and 1,000x more high-performance on-chip high memory than GPU competitors. Cerebras said that the WSE-2 is orders of magnitude larger and more performant than any competing GPU on the market.

“Less than two years ago, Cerebras revolutionized the industry with the introduction of WSE, the world’s first wafer-scale processor,” said Dhiraj Mallik, Vice President Hardware Engineering, Cerebras Systems. “In AI compute, big chips are king, as they process information more quickly, producing answers in less time – and time is the enemy of progress in AI. The WSE-2 solves this major challenge as the industry’s fastest and largest AI processor ever made.”

WSE-1 and CS-1 clients

Cerebras customers include Argonne National Laboratory, Lawrence Livermore National Laboratory, Pittsburgh Supercomputing Center (PSC), EPCC, the supercomputing center at the University of Edinburgh, pharmaceutical leader GlaxoSmithKline (NYSE: GSK), and Tokyo Electron Devices (TYO: 2760).

“The CS-1 allowed us to reduce the experiment turnaround time on our cancer prediction models by 300x over initial estimates, ultimately enabling us to explore questions that previously would have taken years, in mere months,” said Rick Stevens, Argonne National Laboratory Associate Laboratory Director for Computing, Environment, and Life Sciences. “We look forward to seeing what the CS-2 will be able to do with more than double that performance.”

Related Story:  Cerebras Systems Launches AI Chip with a Record 1.2 trillion Transistors