Drilldown: Recipe

DeepSeek V4.1 Flash on 4x 8 GB (Morrowmake) (1) · GLM 5.3 Flash on 2x 8 GB (ExLlamaV3 layer split) (1) · GLM 5.3 Flash on 4x 8 GB (1) · Qwen3.8 Flash Next on 2x 8 GB (Morrowmake EXL3) (1) · Qwen3.8 Flash Next on 3x 8 GB (clee PP3) (1)

DeepSeek-V4.1-Flash-1017 (1) · GLM-5.3-Flash (320B MoE, 18B active) (2) · Qwen3.8-Flash-Next (2)

EXL3 3.05 bpw: turboderp/GLM-5.3-Flash-exl3 (1) · EXL3 5.05 bpw; PLE (n-gram) table read from SSD (1) · Hybrid: routed experts EXL3 2.9 bpw (LibertAIDAI), everything else official FP8/BF16; KV BF16 (1) · W4A16 (INT4 weights, group size 128, FP16 activations): canada-quant/GLM-5.3-Flash-W4A16-MTP (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)

ExLlamaV3 1.5.4 (built for SM80) with a small OpenAI-compatible server (1) · Morrowmake/vllm-cmp170hx (EXL3 path) (1) · Morrowmake/vllm-cmp170hx (vLLM fork, container image) (2) · vLLM nightly 7d0b4e57 plus the author's pure-Python fork clee/vllm-sm80 (branch q38fn/stack) (1)

Flun/glm53-flash-cmp170hx-exl3 at 5933112 (1) · q38fn/stack on nightly-7d0b4e57 (1) · recipe 1.0.0, engine a0c57b5237 (1) · recipe 1.8.0, engine 819146023f (1) · tag qwen38-recipe-1.0.0, engine 18305948ac (1)

8 GB (5)

2 (2) · 3 (1) · 4 (2)

layer split (no tensor parallel), no P2P (1) · PP3, no P2P (works at PP2 on two cards) (1) · TP2 with expert parallel (1) · TP4 (default) or PP4 (1) · TP4 with expert parallel (96 whole experts per card) (1)

1,048,576 (YaRN x4) (1) · 262,144 (3) · 392,960 (profile q8_384k) (1)

60 (1) · 74.4 (1) · 152.6 (1) · 287 (1) · 442.1 (1)

None (1) · 2,651 (1) · 3,161 (1) · 3,818 (1) · 4,831 (1)

64 GiB exposed, --gpu-memory-utilization 0.95 (2) · 64 GiB, --gpu-memory-utilization 0.95 (1) · 64 GiB, gpu-memory-utilization 0.92 (1) · about 58 GiB of 64 GiB (116.6 GiB resident across both) (1)

None (3) · 180 W (2)

None (3) · NVIDIA 580 or newer; P2P needs the Morrowmake/cmpunlocker build (2)

None (3) · cmpunlocker (64 GiB, 74 SMs); Morrowmake/cmpunlocker for P2P (2)

Proven (3) · Reported (2)

nathang6866 · voidgiraffe · HalfVulpes · 1a2b3c0967 · 68x0 · alanoo · anoane · c133 · CurtisAccelerate · darkmaniac7 · dufrelic · greencandle9753 · iIIusi0n · kellertreppe123 · matstatman · mrmannyatx · nestaa51 · pcsv. · plean. · reala7096 · shirogeek · vektor_prime · wgkbfcp · ZackO2o · Zanooda · Other · None

October 10, 2026 (1) · October 16, 2026 (2) · October 20, 2026 (1) · October 24, 2026 (1)

https://github.com/clee/qwen3.8-flash-next-3x170hx (1) · https://github.com/Flun/glm53-flash-cmp170hx-exl3 (1) · https://github.com/Morrowmake/dsv41-flash-cmp170hx-recipe (1) · https://github.com/Morrowmake/glm53-flash-cmp170hx-recipe (1) · https://github.com/Morrowmake/qwen38-flash-next-cmp170hx-recipe (1)