News for hackers. Stuff that matters.

A comment on DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]. Parent comment

Re:DeepSeek open-sources inference optimizations with 60–85%... by alecco on Sunday June 28, 2026 @03:46AM (#48705283)
Isn't GLM-5.2 mostly DeepSeek V3 architecture?

More and more I suspect Z.ai just has deeper pockets and access to the Claude traces while DeepSeek is punching way above their class.

[ Parent | on HN ]
Re:DeepSeek open-sources inference optimizations with 60–85%... by Alifatisk on Sunday June 28, 2026 @01:57PM (#48709741)
Perhaps, but Z.ai contributed with techniques such as IndexShare, which helps reduce computation for larger context windows (1M).
[ Parent | on HN ]
Re:DeepSeek open-sources inference optimizations with 60–85%... by Reubend on Sunday June 28, 2026 @06:10PM (#48712242)
That understates how difficult it is to get to the level of performance they attained. The fact that it surpasses DeepSeek v4 in most ways shows that they accomplished some great work in this space.
[ Parent | on HN ]