Page values for "Qwen3.8 Flash Next on 2x 8 GB (DUFRelic PP2)"
From CMP Wiki
More actions
"Recipe" values
1 row is stored for this page| Field | Field type | Allowed values | Value |
|---|---|---|---|
| page | Page | Qwen3.8 Flash Next on 2x 8 GB (DUFRelic PP2) | |
| model | String | Qwen3.8-Flash-Next | |
| quant | String | W4A16 INT4 g128 with the PLE table repacked INT4 g32: zankich/Qwen3.8-Flash-Next-W4A16-Merlin-INT4PLE-FP8KV; KV BF16 | |
| engine | String | vLLM v0.30.0 + zankich INT4-PLE plugin + six hash-guarded patch layers | |
| engine_version | String | vLLM v0.30.0 (ced6857a); plugin zankich/vllm v0.29.0z (6609bfef) | |
| cards | String | 8 GB · 10 GB · both | 8 GB |
| card_count | Integer | 2 | |
| topology | String | PP2, MTP depth 3 | |
| context | String | 262,144 | |
| decode_tps | Float | 133.39 | |
| prefill_tps | Float | 6,213.66 | |
| vram_per_card | String | 64 GiB, --gpu-memory-utilization 0.90 | |
| power_cap | String | 140 W (200 W observed-draw abort guard) | |
| driver | String | NVIDIA 610.43.02 | |
| unlock_build | String | amoghmunikote/cmpunlocker 6c442eeb (patch 0016 SM reconfig, 74 SMs) plus Bayley P2P patches 0011/0013/0015 | |
| flags | Text | '"`UNIQ--nowiki-0000003D-QINU`"' | |
| status | String | Proven · Reported · Theory · Contested · Warning | Reported |
| tested_by | String | dufrelic | |
| tested_on | Date | 2026-10-03 | |
| source | String | https://github.com/DUFRelic/qwen38-flash-next-recipe | |
| files | Wikitext |