Jump to content
Toggle search
Toggle menu
CMP Wiki
Toggle preferences menu
Toggle personal menu
Not logged in
Your IP address will be publicly visible if you make any edits.

Drilldown: Recipe

More actions
Recipe > card count: 4 & decode tps: 80-150

Use the filters below to narrow your results.

page:
DeepSeek V4.1 Flash on 4x 8 GB (alanoo fidelity build) (1) · DeepSeek V4.1 Flash on 4x 8 GB (CurtisAccelerate TC-MoE) (1) · DeepSeek V4.1 Flash on 4x 8 GB plus RTX PRO 6000 (ExLlamaV3 TP5) (1) · Qwen3.8 Flash Next on 4x 8 GB (1)
model:
DeepSeek-V4.1-Flash-1017 (3) · Qwen3.8-Flash-Next (1)
quant:
AWQ W4A16: wtdcode/Qwen3.8-Flash-Next-AWQ-W4A16 (1) · EXL3 3.5 bpw (LibertAIDAI; routed experts 4 bpw on layers 0-9 and 30-39, 3 bpw on 10-29; attention and shared experts 6 bpw) plus the official Engram shards (1) · Hybrid: all 40 expert layers at 3 bits (LibertAI 2.9 bpw pack with layers 18-22 grafted from the 3.5 bpw pack), official FP8 dense tensors (1) · Hybrid: LibertAI EXL3 2.9 bpw routed experts (layers 0-39), official FP8/FP4 for attention, shared experts, head, draft and Engram; KV fp8_ds_mla (1)
engine:
ExLlamaV3 fork (anoane/exllamav3-anemone) + TabbyAPI fork (anoane/tabbyAPI) (1) · GNmac-lab recipe image 1.2.0 (vLLM with sm_80 patches) plus three patches (1) · Morrowmake/vllm-cmp170hx (bind-mounted sm80 package) plus the author's tensor-core MoE kernels (1) · wtdcode/vllm-backport (image lazymio/vllm-backport:latest-sm80) (1)
engine version:
GNmac-lab at 2ad5741ab1; image ds41-cmp170hx:1.2.0-fidelity (1) · latest-sm80 tag as of 2026-10-16 (1) · tag recipe-09 (309ef6d1); TabbyAPI deepseek-v41 (ce250ff) (1) · vllm-cmp170hx 18305948ac; TC-MoE fb9786ec (1)
cards:
8 GB (4)
card count:
4 (4)
topology:
TP4 with expert parallel, MTP 3 (1) · TP4, P2P on (1) · TP4, peer access off (NCCL_P2P_DISABLE=1), host all-reduce (1) · TP5 across four CMP 170HX and one RTX PRO 6000 Blackwell (native backend, reduces through host memory); MTP draft on the PRO (1)
context:
1,000,000 (YaRN x4 from 262,144) (1) · 1,048,576 FP16 (1) · 262,144 (1) · 262,144 (weights address 1,048,576) (1)
decode tps:
99.9 (1) · 105 (1) · 116.49 (1) · 144.2 (1)
prefill tps:
1,523 (1) · 2,101.4 (1) · 2,116 (1) · 2,514 (1)
vram per card:
64 GB (1) · 64 GiB (64,911 MiB usable), GPU_MEMORY_UTILIZATION 0.98 (1) · about 62.5 GB used of 64 GB, --gpu-memory-utilization 0.97 (1) · gpu_split 58 GB per CMP, 70 GB on the PRO (1)
power cap:
None (1) · 140 W (1) · 150 W per CMP (PRO 400 W) (1) · 200 W (1)
driver:
None (2) · NVIDIA open 610.43.02 from a cmpunlocker fork with a P2P patch (1) · NVIDIA open 610.57.04 patched with Morrowmake/cmpunlocker 1f2c498 (profile 8gb) (1)
unlock build:
None (2) · cmpunlocker fork with P2P (without Morrowmake's TRAP31 patch) (1) · Morrowmake/cmpunlocker 1f2c498; modprobe RmForceEnableGen2=1, RMPcieLinkSpeed=0x1, ForceP2P=0x11 (1)
status:
Reported (4)
tested by:
alanoo (1) · anoane (1) · CurtisAccelerate (1) · matstatman (1)
tested on:
October 16, 2026 (1) · October 20, 2026 (1) · October 22, 2026 (1) · October 27, 2026 (1)
source:
https://discord.com/channels/1518634379604136178/1528826330454429706/1560602001908170753 (1) · https://github.com/alanoo81/deepseek-v41-flash-4x-cmp170hx (1) · https://github.com/anoane/09-deepseek-v41-flash-1017-exl3-tp5-1m (1) · https://github.com/CurtisAccelerate/DeepSeekv41-Flash-1017-EXL3-CMP170HX (1)

Showing below up to 4 results in range #1 to #4.

View (previous 250 | next 250) (20 | 50 | 100 | 250 | 500)


View (previous 250 | next 250) (20 | 50 | 100 | 250 | 500)

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