Skip to content
Discussion options

You must be logged in to vote

Flash-Decoding parallelizes the attention reduction operation across the sequence length dimension in addition to batch and head dimensions. By splitting the KV cache of a long prompt across multiple GPU thread blocks, it maximizes memory bus utilization during single-token generation.

Replies: 1 comment

Comment options

Ultron09
Aug 27, 2026
Maintainer Author

You must be logged in to vote
0 replies
Answer selected by Ultron09
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
1 participant