Drilldown: Recipe - CMP Wiki

Jump to content

Choose a table:

Recipe > page

: Qwen3.8 27B on 1x 8 GB (W4A16 vLLM)

or Qwen3.8 Flash Next on 2x 8 GB (Plean PP2)

or Qwen3.8 Flash Next on 3x 8 GB (clee PP3)

Use the filters below to narrow your results.

Qwen3.8-27B (dense) (1) · Qwen3.8-Flash-Next (2)

Mixed INT4 AutoRound (Minachist/Qwen3.8-Flash-Next-INT4-Mixed-AutoRound, about 65 GB) with the full BF16 PLE table on NVMe (1) · W4A16 (INT4 weights, Marlin kernels) (1) · W4A16 AutoRound: Intel/Qwen3.8-Flash-Next-W4A16-AutoRound; 95.4 GiB BF16 PLE table streamed from NVMe; KV fp8 e4m3 (software-decoded on SM80) (1)

vLLM (1) · vLLM 0.29.1rc1 nightly plus the checkpoint author's patches (1) · vLLM nightly 7d0b4e57 plus the author's pure-Python fork clee/vllm-sm80 (branch q38fn/stack) (1)

None (2) · q38fn/stack on nightly-7d0b4e57 (1)

8 GB (3)

1 (1) · 2 (1) · 3 (1)

PP2, MTP 3, no P2P (1) · PP3, no P2P (works at PP2 on two cards) (1) · single card (1)

1,048,576 (YaRN x4) (1) · 262,144 (256K in the post) (1) · 262,144 native (YaRN to about 500k in the recipe's long variant) (1)

74.4 (1) · 170 (1) · 197 (1)

None (1) · 4,831 (1) · 5,800 (1)

64 GB (2) · 64 GiB, gpu-memory-utilization 0.92 (1)

None (3)

None (3)

None (3)

Proven (1) · Reported (2)

c133 (1) · kellertreppe123 (1) · plean. (1)

October 5, 2026 (1) · October 7, 2026 (1) · October 10, 2026 (1)

https://github.com/clee/qwen3.8-flash-next-3x170hx (1) · https://www.bozocentral.com/2x-cmp170hx-qwen3-flash-next (1) · https://www.reddit.com/r/LocalLLM/comments/1w2f03e/qwen3827b_w4a16_on_one_64_gb_cmp_170hx_16070/ (1)

Showing below up to 3 results in range #1 to #3.

Q

Qwen3.8 27B on 1x 8 GB (W4A16 vLLM)

Qwen3.8 Flash Next on 2x 8 GB (Plean PP2)

Qwen3.8 Flash Next on 3x 8 GB (clee PP3)

Retrieved from "https://cmp.wiki/wiki/Special:Drilldown/Recipe"