biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Kernel] Add FP8 KV cache support to Triton MLA decode attention (#34597)

Browse Source 
Signed-off-by: grimulkan <grimulkan@gmail.com>
This commit is contained in:
grimulkan
2026-03-12 10:32:34 -05:00
committed by GitHub
parent abcffbba8c
commit a1257fd1ea
4 changed files with 192 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/design/attention_backends.md
View File

@@ -213,5 +213,5 @@ configuration.
| `ROCM_AITER_MLA` | fp16, bf16 | `auto`, `bfloat16`, `fp8`, `fp8_e4m3`, `fp8_e5m2` | 1 | Any | ❌ | ❌ | ❌ | ❌ | Decoder | N/A |
| `ROCM_AITER_MLA_SPARSE` | fp16, bf16 | `auto`, `bfloat16` | 1 | Any | ❌ | ✅ | ❌ | ❌ | Decoder | N/A |
| `ROCM_AITER_TRITON_MLA` | fp16, bf16 | `auto` | Any | Any | ❌ | ❌ | ❌ | ❌ | Decoder | N/A |
| `TRITON_MLA` | fp16, bf16 | `auto`, `bfloat16` | %16 | Any | ❌ | ❌ | ❌ | ✅ | Decoder | Any |
| `TRITON_MLA` | fp16, bf16 | `auto`, `bfloat16`, `fp8`, `fp8_e4m3` | %16 | Any | ❌ | ❌ | ❌ | ✅ | Decoder | Any |
| `XPU_MLA_SPARSE` | fp16, bf16 | `auto`, `bfloat16` | Any | 576 | ❌ | ✅ | ❌ | ❌ | Decoder | Any |