Cerebras and AMD partner to build the world’s fastest disaggregated AI inference solution
SiliconANGLE Thomas Godwin ● Covered by 4 sources
Cerebras and AMD are teaming up to split AI inference into two specialized jobs instead of one. Together they claim 5x more speed per watt than current setups, which could mean much cheaper AI answers.
There's a dirty secret in AI inference: the two halves of generating a response don't actually want the same kind of hardware. The
My take
placeholder
Read more about this at: SiliconANGLE
