Drilldown: Recipe - CMP Wiki

Jump to content

Choose a table:

Recipe > prefill tps: None

Use the filters below to narrow your results.

DeepSeek V4.1 Flash on 2x 8 GB (Project Maya) (1) · DeepSeek V4.1 Flash on 6x 8 GB (TP6) (1) · GLM 5.3 Flash on 2x 8 GB (ExLlamaV3 layer split) (1) · GLM 5.3 Flash on 3x 8 GB (1) · GLM 5.3 on 8x 8 GB (Int4-Int8Mix PP8) (1) · Qwen3.8 27B on 1x 8 GB (W4A16 vLLM) (1) · Qwen3.8 Flash Next on 2x 8 GB (vektorprime PP2) (1)

DeepSeek-V4.1-Flash-1017 (1) · DeepSeek-V4.1-Flash-1017 (748B in the Maya README) (1) · GLM-5.3 (full 78-layer, not Flash) (1) · GLM-5.3-Flash (320B MoE, 18B active) (2) · Qwen3.8-27B (dense) (1) · Qwen3.8-Flash-Next (1)

AWQ INT4: cyankiwi/Qwen3.8-Flash-Next-AWQ-INT4; 95 GiB BF16 PLE table in host RAM (FP8 table optional) (1) · EXL3 3.05 bpw: turboderp/GLM-5.3-Flash-exl3 (1) · GLM-5.3-Int4-Int8Mix (INT4 / INT8 mix); KV fp8_ds_mla (1) · Maya GGUF (Maya-S 340 GB or Maya-XS 277 GB; Engram tables stay on SSD) (1) · Official checkpoint: dense MXFP8 through an SM80 Marlin W8A16 path, routed experts MXFP4 through W4A16; KV FP8 (1) · W4A16 (INT4 weights, Marlin kernels) (1) · W4A16, about 4.3 bpw, 163 GiB on disk (1)

Custom SM80 vLLM (patched from public commit 79a7108d9) (1) · ExLlamaV3 1.5.4 (built for SM80) with a small OpenAI-compatible server (1) · Morrowmake/vllm-cmp170hx (via the glm53-flash-cmp170hx-recipe) (1) · Project Maya (Strata and llama.cpp lineage) (1) · vLLM (1) · vLLM (build not stated) (1) · vLLM nightly + PR #53899 (a5530b9) + the repo's patch scripts (1)

None (3) · 0.1.dev1+ga5530b90c.d20260829 (1) · Flun/glm53-flash-cmp170hx-exl3 at 5933112 (1) · not stated (1) · reports 0.0.0; patches/public-base-to-runtime.patch (109 files) (1)

8 GB (7)

1 (1) · 2 (3) · 3 (1) · 6 (1) · 8 (1)

engine-managed expert tiering across VRAM, host RAM and NVMe (1) · layer split (no tensor parallel), no P2P (1) · PP2 with PLE CPU offload (1) · PP3 (1) · PP8, TP1 (1) · single card (1) · TP6 (virtual-head padding), PP1 (1)

None (1) · 1,048,576 (2) · 262,144 (1) · 262,144 native (YaRN to about 500k in the recipe's long variant) (1) · 392,960 (profile q8_384k) (1) · about 1,000,000 (8-bit KV cache) (1)

None (1) · 0 - 200 (6)

64 GB (2) · 64 GB, --gpu-memory-utilization 0.95 (1) · 64 GiB (1) · 64 GiB, --gpu-memory-utilization 0.97 (1) · 64 GiB; weights PP0 39.48 GiB, PP1 45.53 GiB; --gpu-memory-utilization 0.9 (1) · about 58 GiB of 64 GiB (116.6 GiB resident across both) (1)

None (6) · 200 W (1)

None (7)

None (7)

Proven (1) · Reported (6)

1a2b3c0967 (1) · 68x0 (1) · kellertreppe123 (1) · pcsv. (1) · shirogeek (1) · vektor_prime (1) · wgkbfcp (1)

August 2026 (1) · October 2026 (6)

https://discord.com/channels/1518634379604136178/1518695289278566450/1561743826945843304 (1) · https://discord.com/channels/1518634379604136178/1528826330454429706/1557523316120748073 (1) · https://github.com/Flun/glm53-flash-cmp170hx-exl3 (1) · https://github.com/mw00/project-maya#how-fast-is-it (1) · https://github.com/vektorprime/qwen38-flash-next-pp2 (1) · https://github.com/wgkbfcp/deepseek-v4.1-flash-1017-cmp170hx-recipe (1) · https://www.reddit.com/r/LocalLLM/comments/1w2f03e/qwen3827b_w4a16_on_one_64_gb_cmp_170hx_16070/ (1)

Showing below up to 7 results in range #1 to #7.

D

G

GLM 5.3 Flash on 2x 8 GB (ExLlamaV3 layer split)

Q

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

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

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