Drilldown: Recipe

GLM 5.3 Flash on 3x 8 GB (1) · Qwen3.8 Flash Next on 2x 8 GB (ElectroHobo W4A16 TP2) (1) · Qwen3.8 Flash Next on 3x 8 GB (clee PP3) (1)

GLM-5.3-Flash (320B MoE, 18B active) (1) · Qwen3.8-Flash-Next (1) · Qwen3.8-Flash-Next (125B MoE per the recipe) (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) · W4A16, about 4.3 bpw, 163 GiB on disk (1) · W4A16: halt95/Qwen3.8-Flash-Next-W4A16-Merlin (AutoRound INT4 g128 experts, INT8 GDN projections, FP8 n-gram table) (1)

Morrowmake/vllm-cmp170hx (via the glm53-flash-cmp170hx-recipe) (1) · Morrowmake/vllm-cmp170hx image (Qwen EXL3 release) with patch files mounted read-only (1) · vLLM nightly 7d0b4e57 plus the author's pure-Python fork clee/vllm-sm80 (branch q38fn/stack) (1)

None · recipe 0.2.1; engine 18305948ac · 0.1.dev1+ga5530b90c.d20260829 · 0.29.1rc1.dev402+ga5a30471f.ple1 · Deepinfer-SM80 2a828fc lineage; vLLM v0.28.1rc0 base · Deepinfer-SM80, vLLM v0.28.1rc0 base · Flun/glm53-flash-cmp170hx-exl3 at 5933112 · GNmac-lab at 2ad5741ab1; image ds41-cmp170hx:1.2.0-fidelity · latest-sm80 tag as of 2026-10-16 · not stated · q38fn/stack on nightly-7d0b4e57 · recipe 0.3.1; engine 18305948ac · recipe 1.0.0 (commit d4d1f525), engine a0c57b52 · recipe 1.0.0, engine a0c57b5237 · recipe 1.2.0 · recipe 1.8.0, engine 819146023f · recipe 31318fd; patches 0001-0020 · reports 0.0.0; patches/public-base-to-runtime.patch (109 files) · tag qwen38-recipe-1.0.0, engine 18305948ac · tag recipe-09 (309ef6d1); TabbyAPI deepseek-v41 (ce250ff) · upstream dev 0f685aa9fd plus 9 patches; repo db01c256 · v1.0.6 (this row) · vLLM 0.30.1rc1, fork commit 378c37b0 · vLLM v0.30.0 (ced6857a); plugin zankich/vllm v0.29.0z (6609bfef) · vllm-backport 3142ded1 · vllm-cmp170hx 18305948ac; TC-MoE fb9786ec · Other

8 GB (3)

2 (1) · 3 (2)

PP3 (1) · PP3, no P2P (works at PP2 on two cards) (1) · TP2 with whole-expert expert parallel (1)

1,048,576 (YaRN x4) (1) · 262,144 (1) · about 1,000,000 (8-bit KV cache) (1)

None (1) · 74.4 (1) · 349.7 (1)

None (1) · 4,696 (1) · 4,831 (1)

64 GiB (1) · 64 GiB, gpu-memory-utilization 0.92 (1) · 64 GiB, GPU_UTIL 0.95 (fixed) (1)

None (2) · 180 W (1)

None (2) · 580 or newer (measured on 610.57.04); open modules with HMM for the table-on-disk mode (1)

None (2) · cmpunlocker (64 GiB, 74 SMs) with a P2P-capable build (1)

Proven (2) · Reported (1)

c133 (1) · nathang6866 (1) · shirogeek (1)

October 10, 2026 (1) · October 19, 2026 (1) · October 23, 2026 (1)

https://discord.com/channels/1518634379604136178/1518695289278566450/1561743826945843304 (1) · https://github.com/clee/qwen3.8-flash-next-3x170hx (1) · https://github.com/ElectroHobo/qwen38-flash-next-w4a16-2x-cmp170hx-recipe (1)