I'm not sure that I buy their conclusion that more compute during inference is good.
Yes, batch=1 inference is mostly memory bandwidth bound, not GPU compute bound. But no provider does batch=1 inference. Everyone groups all the requests into a batch, and the GPU computes them together.
With a fused kernel, that means the GPU streams the tensors from VRAM, and does a bunch of compute on different conversations in the batch, at the same time.
If they increase the amount of compute required per token, that just reduces the maximum batch size a GPU can handle. In practice, yes this does mean each GPU can serve less users. Providers aren't leaving GPU cores idle normally during inference.
Yes, batch=1 inference is mostly memory bandwidth bound, not GPU compute bound. But no provider does batch=1 inference. Everyone groups all the requests into a batch, and the GPU computes them together.
With a fused kernel, that means the GPU streams the tensors from VRAM, and does a bunch of compute on different conversations in the batch, at the same time.
If they increase the amount of compute required per token, that just reduces the maximum batch size a GPU can handle. In practice, yes this does mean each GPU can serve less users. Providers aren't leaving GPU cores idle normally during inference.