Re:Intel Announces Aurora GenAI, ChatGPT Competitor with 1T ...
by HybridCurve on Monday May 22, 2023 @07:18PM
(#36037929)
Aurora Supercomputer:
- 200Gb dragonfly interconnect fabric
- 10,000 nodes with 2xCPU + 6xGPU with unified memory
- 10PB of agreggate RAM
- 230PB of storage
Scale matters in supercomputing. Keeping latency low between everything allows you to solve different classes of problems that gets no speed up benefit by simply offloading compute to nvidia cards. It makes a huge difference how everything is connected.
- 200Gb dragonfly interconnect fabric
- 10,000 nodes with 2xCPU + 6xGPU with unified memory
- 10PB of agreggate RAM
- 230PB of storage
Scale matters in supercomputing. Keeping latency low between everything allows you to solve different classes of problems that gets no speed up benefit by simply offloading compute to nvidia cards. It makes a huge difference how everything is connected.