Drilldown: Recipe - CMP Wiki

Jump to content

Choose a table:

Recipe > page

: DeepSeek V4.1 Flash on 4x 8 GB (Morrowmake)

or GLM 5.3 Flash on 2x 8 GB (ExLlamaV3 TP2)

Use the filters below to narrow your results.

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

EXL3 3.05 bpw: turboderp/GLM-5.3-Flash-exl3; DFlash2 drafter converted to EXL3 6 bpw (1) · Hybrid: routed experts EXL3 2.9 bpw (LibertAIDAI), everything else official FP8/BF16; KV BF16 (1)

ExLlamaV3 (patched, pinned) (1) · Morrowmake/vllm-cmp170hx (vLLM fork, container image) (1)

recipe 1.0.0, engine a0c57b5237 (1) · upstream dev 0f685aa9fd plus 9 patches; repo db01c256 (1)

8 GB (2)

TP2 over NCCL (GPU P2P), DFlash2 K7 plus hybrid drafting (1) · TP4 with expert parallel (96 whole experts per card) (1)

262,144 (1) · 392,960 per request; 655,360-token Q8 KV pool shared by two requests (1)

107.67 (1) · 287 (1)

1,533 (1) · 3,818 (1)

64 GiB exposed, --gpu-memory-utilization 0.95 (1) · about 62.9 / 62.3 GiB idle, peaks about 64.4 / 63.0 GiB with vision (1)

180 W (1) · 180 W (author's cards); about 170 W with a modest memory OC in his Discord numbers (1)

610.57.04 (1) · NVIDIA 580 or newer; P2P needs the Morrowmake/cmpunlocker build (1)

cmpunlocker (64 GiB); P2P-capable driver for the NCCL backend (1) · cmpunlocker (64 GiB, 74 SMs); Morrowmake/cmpunlocker for P2P (1)

Reported (2)

nestaa51 (1) · voidgiraffe (1)

October 13, 2026 (1) · October 24, 2026 (1)

https://github.com/alexvnesta/glm53-flash-cmp170hx-tp (1) · https://github.com/Morrowmake/dsv41-flash-cmp170hx-recipe (1)

Showing below up to 2 results in range #1 to #2.

D

DeepSeek V4.1 Flash on 4x 8 GB (Morrowmake)

G

GLM 5.3 Flash on 2x 8 GB (ExLlamaV3 TP2)

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