From EE Times, July 20:
G42 and Cerebras have partnered to build a significant AI supercomputer, Condor Galaxy 1, to be built this year in nearby Santa Clara, based on Cerebras hardware. Condor Galaxy 1, which will consist of 64 connected Cerebras CS-2 wafer-scale AI accelerators, represents a deal worth “in excess of $100 million,” Cerebras CEO Andrew Feldman told EE Times. Following Condor Galaxy 1, strategic partners Cerebras and G42, an Abu Dhabi–based technology group, plan eight further 64-CS-2 clusters inside and outside the United States, which are set to reach a total of 576 connected CS-2 systems by the end of next year.
The total cost is expected to be in excess of $900 million.
“G42 was looking for a partner that had expertise in building, managing and operating supercomputers, and there are not so many of us anymore who are implementing massive generative AI models and manipulating, cleaning and managing big datasets, who could deliver infrastructure with fast compute that’s easy to use, flexible and had a track record of exceptional collaboration,” Feldman said. “Those were some of the reasons they chose us.”
Condor Galaxy 1 represents the first supercomputer-scale installation of AI accelerator hardware from any startup, anywhere, and validates the market for specialized AI hardware outside of GPUs, a sector that has attracted billions of dollars of venture capital over the last five years.
Describing the deal as “a challenge on multiple fronts,” Feldman said, “It’s a challenge to Nvidia. It’s a challenge to hyperscalers. It’s saying, ‘Look, you don’t need to be one of these [big] guys to stand up extraordinary compute.’”....
....MUCH MORE
Previously on Cerebras:January 2020
"Cerebras’ Giant Chip Will Smash Deep Learning’s Speed Barrier"
From IEEE Spectrum:
Computers using Cerebras’s chip will train these AI systems in hours instead of weeks
That's a big chip.
May 2018
"Artificial intelligence chips are a hot market." (NVDA; GOOG)
August 2019
Artificial Intelligence: The World's Largest Chip Is As Big As Your Head and Contains 1.2 Trillion Transistors
August 2019
Chips: "Two contrasting approaches to AI chips emerged at Hot Chips Conference..." (AMD; NVDA; XLNX)
September 2019
Chips: The World's Largest Semiconductor Chip Will Power "Supercomputing hog U.S. Department of Energy"
A Mini-Masters Class On What's What In AI Chips, Machine Learning and Programming
And many more, use the 'search blog' box if interested.
