41554 blogs · 4,417,066 posts Claim your blog

Record · corpus.blog/posts/01a0876e-301c-71ea-966c-d1995366d008

2x GH200 for LLM inference, Part 3: GLM-5.2, expert offload, and the CPU question

dnhkng.github.io · published 16 June 2026

A record is what was published and who pointed at it. The text of the post is not held here: read it at the source, or at its Wayback capture.

Read the post at dnhkng.github.io Wayback capture

0blogs citing

First cited: —
Most recent: —
Rank this month: not ranked
Links from this post: 6
Words captured: 3,954

Cited by

No blog we hold has cited this post yet.

Links from this post

Link

Host

https://dnhkng.github.io/posts/gh200-benchmarking/

dnhkng.github.io

https://dnhkng.github.io/posts/gh200-benchmarking-part-2/

dnhkng.github.io

https://huggingface.co/zai-org/GLM-5.2-FP8

huggingface.co

https://huggingface.co/cyankiwi/GLM-5.2-AWQ-INT4

huggingface.co

https://huggingface.co/dnhkng/GLM-5.2-AWQ-INT4-FP8-MTP-delta

huggingface.co

2x GH200 for LLM inference, Part 4: DeepSeek V4 Flash - SGLang vs vLLM at 1M context Read at the source ↗

dnhkng.github.io