In your forward pass section you give a lot of emphasis to FlashAttention, but it might be worth mentioning Paged Attention as well (which was the paper written by the vLLM authors and I believe was the genesis of the project). PA-style block tables are now supported in most fused attention kernels, but vLLM originally came up with it and it's the main reason why vLLM has such high throughput!

Thanks for writing the article!

I didn't quite get

Note that during the prefill phase, all prompt tokens from a request can be processed in one batch. This is possible because the query (Q) tensors, calculated from the tokens immediately before them, are available for each prompt token position.

I know that in practice prefill is much faster than inference. Would watching the 2h video from Karpathy help me understand why?

Great write up, it would be interesting to see a lot of those covered features in comparison to other frameworks!

Thanks for this! Learnt a lot.

Curious to understand how do we ensure that the same model instance gets requests from the same client/user? Since conversations are stateful and the model needs context from previous turns of the conversation.

Is this happening at the load balancer layer?

Batching in vLLM doesn't combine prompts into the same context - it processes separate requests in parallel while sharing compute resources, so there's no perplexity tradeoff, just efficiency gains.

I would like to know what inference speeds they are achieving exactly on what hardware. I skimmed and searched the article and didn't find that info.