Drilldown: Recipe

DeepSeek V4.1 Flash on 6x 8 GB (Deepinfer-SM80 TP2 x PP3) (1) · Qwen3.8 Flash Next on 2x 8 GB (Plean PP2) (1) · Qwen3.8 Flash Next on 2x 8 GB (Qwen-Flash-SM80, GDS TEP2) (1)

DeepSeek-V4.1-Flash-1017 (1) · Qwen3.8-Flash-Next (1) · Qwen3.8-Flash-Next (nvidia/Qwen3.8-Flash-Next-NVFP4, revision fc694b54, 11 shards, 132.68 GB) (1)

Mixed INT4 AutoRound (Minachist/Qwen3.8-Flash-Next-INT4-Mixed-AutoRound, about 65 GB) with the full BF16 PLE table on NVMe (1) · NVFP4 main weights through Marlin (no native FP4 on sm_80), mixed FP8 and BF16 MTP weights, FP8 PLE table (about 51.2 GB) kept on NVMe (1) · Official checkpoint (FP8 dense, FP4 experts); software FP8 on sm_80 (1)

Qwen-Flash-SM80 overlay on vllm/vllm-openai:qwen38-flash-next (digest-pinned), FlashInfer PCIe IPC AllReduce, GDS/cuFile over NVMe P2PDMA (1) · vLLM 0.29.1rc1 nightly plus the checkpoint author's patches (1) · vLLM PR #56201 head (79a7108d) plus the two Deepinfer-SM80 patches and 30 drop-in modules (1)

None (1) · Deepinfer-SM80 2a828fc lineage; vLLM v0.28.1rc0 base (1) · v0.3.0 (commit 4d229f1, 2026-10-14); base vLLM 0.1.dev20073+g8e685d198 (1)

8 GB (3)

2 (2) · 6 (1)

PP2, MTP 3, no P2P (1) · TEP2 (TP2 plus expert parallel), MTP 6; one card at PCIe Gen2 x16 (x16 mod), one at Gen2 x4; GPU P2P required (1) · TP2 x PP3, partition 14,14,12 (1)

1,048,576 (1) · 262,144 (1) · 262,144 (256K in the post) (1)

160.8 (1) · 168.29 (1) · 197 (1)

1,986.38 (1) · 4,037 (1) · 5,800 (1)

64 GB (1) · 64 GB, --gpu-memory-utilization 0.93 (v0.3.0) (1) · 64 GiB (1)

None · 180 W · 200 W · 100 W · 140 W · 140 W (200 W observed-draw abort guard) · 150 W per CMP (PRO 400 W) · 170 W (author's 10-card run) · 180 W (author's cards); about 170 W with a modest memory OC in his Discord numbers · 180 W (nvidia-smi -pl 180) · 180 W (one card unstable at 250 W) · Other

None (2) · NVIDIA open 610.43.03, CUDA 13.0 (1)

None (2) · bayley/cmpunlocker 5a7bb4b, sudo ./install.sh --p2p --no-iommu (1)

Reported · Proven · Other · None

HalfVulpes (1) · nguyenthimy2022kg-alt (1) · plean. (1)

September 8, 2026 (1) · October 7, 2026 (1) · October 28, 2026 (1)

https://github.com/HalfVulpes/Deepinfer-SM80 (1) · https://github.com/nguyenthimy2022kg-alt/Qwen-Flash-SM80-170HX (1) · https://www.bozocentral.com/2x-cmp170hx-qwen3-flash-next (1)