Page values for "Qwen3.8 Flash Next on 4x 8 GB"
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 4x 8 GB | |
| model | String | Qwen3.8-Flash-Next | |
| quant | String | AWQ W4A16: wtdcode/Qwen3.8-Flash-Next-AWQ-W4A16 | |
| engine | String | wtdcode/vllm-backport (image lazymio/vllm-backport:latest-sm80) | |
| engine_version | String | latest-sm80 tag as of 2026-10-16 | |
| cards | String | 8 GB · 10 GB · both | 8 GB |
| card_count | Integer | 4 | |
| topology | String | TP4 with expert parallel, MTP 3 | |
| context | String | 1,000,000 (YaRN x4 from 262,144) | |
| decode_tps | Float | 116.49 | |
| prefill_tps | Float | 2,101.4 | |
| vram_per_card | String | 64 GB | |
| power_cap | String | ||
| driver | String | ||
| unlock_build | String | ||
| flags | Text | --tensor-parallel-size 4 --enable-expert-parallel --speculative-config mtp 3 --max-model-len 1000000 --max-num-seqs 8 --max-num-batched-tokens 2048 cudagraph FULL_AND_PIECEWISE --prefix-match-unit 16 --mamba-cache-mode align; VLLM_PLE_CPU_OFFLOAD=0 | |
| status | String | Proven · Reported · Theory · Contested · Warning | Reported |
| tested_by | String | matstatman | |
| tested_on | Date | 2026-10-16 | |
| source | String | https://discord.com/channels/1518634379604136178/1528826330454429706/1560602001908170753 | |
| files | Wikitext |