Sign In

Communications of the ACM

ACM News

Cerebras' New Monster AI Chip Adds 1.4 Trillion Transistors


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
Cerebras CEO Andrew Feldman and the Wafer Scale Engine 2.

Cerebras Systems' new Wafer Scale Engine 2 is just as big physically as its predecessor, the largest single computer chip ever built, but has enormously increased amounts of, well, everything.

Credit: Peter Adams

Almost from the moment Cerebras Systems announced a computer based on the largest single computer chip ever built, the Silicon Valley startup declared its intentions to build an even heftier processor. Today, the company announced that its next-gen chip, the Wafer Scale Engine 2 (WSE 2), will be available in the 3rd quarter of this year. WSE 2 is just as big physically as its predecessor, but it has enormously increased amounts of, well, everything. The goal is to keep ahead of the ever-increasing size of neural networks used in machine learning.

"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," Dhiraj Malik, vice president of hardware engineering said in a statement. 

Cerebras has always been about taking a logical solution to the problem of machine learning to the extreme. Training neural networks takes too long—weeks for the big ones when Andrew Feldman cofounded the company in 2015. The biggest bottleneck was that data had to shuttle back and forth between the processor and external DRAM memory, eating up both time and energy. The inventors of the original Wafer Scale Engine figured that the answer was to make the chip big enough to hold all the data it needed right alongside its AI processor cores. With gigantic networks for natural language processing, image recognition, and other tasks on the horizon, you'd need a really big chip. How big? As big as possible, meaning the size of an entire wafer of silicon (with the round bits cut off), or 46,225 square millimeters.

From IEEE Spectrum
View Full Article

 


 

No entries found