LLM inference is a two-phase process. The first phase is prompt processing aka prefill. It's compute-heavy but requires relatively low memory bandwidth. The second phase is token generation aka decode, which doesn't require much in the way of FLOPs but wants as much memory bandwidth as possible.
This announcement is for a system to do the first phase on Helios and the second phase on WSE.
Nobody knows the answer to that, but if it’s efficiency they are aiming for, which results in lower inference costs, it’ll last for a while. The inference to training ratio will only get higher and the margins will only get lower.
I'm not sure I understand. It's Helios but with WSE attached? And it uses one or the other depending on some criteria?
LLM inference is a two-phase process. The first phase is prompt processing aka prefill. It's compute-heavy but requires relatively low memory bandwidth. The second phase is token generation aka decode, which doesn't require much in the way of FLOPs but wants as much memory bandwidth as possible.
This announcement is for a system to do the first phase on Helios and the second phase on WSE.
It’s disaggregated so you Know it’s good
The two major sides of the system excel at different things that happen to be complementary in AI inference.
They're a bit late to the party.
Why would you say that?
lol when is the music gonna stop?
Nobody knows the answer to that, but if it’s efficiency they are aiming for, which results in lower inference costs, it’ll last for a while. The inference to training ratio will only get higher and the margins will only get lower.
When Etched becomes GA.