[minimaxir]

A few days ago I was just thinking that Google never talked about their diffusion text generation model after demoing it at I/O a year ago. The rumor is that it was too expensive to run, but with the provided chart using the same 1x H100 hardware and comparing DiffusionGemma to regular Gemma, that shouldn't be the case. I'm curious what the downside for this speed is here aside from being slightly weaker than Gemma.

[ac29]

> I'm curious what the downside for this speed is here

"DiffusionGemma's speedup is designed for local and low-concurrency inference. In high-QPS cloud serving, autoregressive models can be deployed to saturate compute efficiently, so DiffusionGemma's parallel decoding offers diminishing returns and can result in higher serving costs"

[GaggiX]

Well with a standard autoregressive model you can generate for example 256 tokens at once if you have 256 users, with this approach you can generate 256 tokens for a single user but you need several forward steps.

So the diffusion process takes more GFLOPs, if you have enough users you can already balance memory and compute.

[minimaxir]

Batching is a fair counterpoint.