The letter

One letter a week, in your inbox.

The week's signal, what shipped, what is worth running tonight, and the editor's note. No tracking, no ads, nothing else.

We keep your address, your language and the date you joined, nothing else. Every letter has a one-click unsubscribe link that deletes the record.

← Accept All   Archive
Zenn (AI)Japan

GLM 5.3 Flash (NVFP4) を DGX Spark 2 台で動かす

September 20

NVIDIA DGX Spark(GB10)2台を200Gbps QSFP(RoCEv2)で直結し、LibertAIDAI/GLM-5.3-Flash-NVFP4(マルチモーダル image+video 対応、NVFP4 量子化の 320B級 MoE モデル)を分散推論(Tensor Parallelism = 2)でデプロイした際の手順をまとめました。

Read at Zenn (AI) ↗

More from Zenn (AI) on Accept All

More from Zenn (AI) on Accept All.