|
1 | 1 | # Release note |
2 | 2 |
|
| 3 | +## v0.10.0rc1 - 2025.08.07 |
| 4 | + |
| 5 | +This is the 1st release candidate of v0.10.0 for vLLM Ascend. Please follow the [official doc](https://vllm-ascend.readthedocs.io/en/) to get started. V0 is completely removed from this version. |
| 6 | + |
| 7 | +### Highlights |
| 8 | +* Disaggregate prefill works with V1 engine now. You can take a try with DeepSeek model [#950](https://github.com/vllm-project/vllm-ascend/pull/950), following this [tutorial](https://github.com/vllm-project/vllm-ascend/blob/main/examples/disaggregated_prefill_v1/README.md). |
| 9 | +* W4A8 quantization method is supported for dense and MoE model now. [#2060](https://github.com/vllm-project/vllm-ascend/pull/2060) [#2172](https://github.com/vllm-project/vllm-ascend/pull/2172) |
| 10 | + |
| 11 | +### Core |
| 12 | +* Ascend PyTorch adapter (torch_npu) has been upgraded to `2.7.1.dev20250724`. [#1562](https://github.com/vllm-project/vllm-ascend/pull/1562) And CANN hase been upgraded to `8.2.RC1`. [#1653](https://github.com/vllm-project/vllm-ascend/pull/1653) Don’t forget to update them in your environment or using the latest images. |
| 13 | +* vLLM Ascend works on Atlas 800I A3 now, and the image on A3 will be released from this version on. [#1582](https://github.com/vllm-project/vllm-ascend/pull/1582) |
| 14 | +* Kimi-K2 with w8a8 quantization, Qwen3-Coder and GLM-4.5 is supported in vLLM Ascend, please following this [tutorial](https://vllm-ascend.readthedocs.io/en/latest/tutorials/multi_node_kimi.md.html) to have a try. [#2162](https://github.com/vllm-project/vllm-ascend/pull/2162) |
| 15 | +* Pipeline Parallelism is supported in V1 now. [#1800](https://github.com/vllm-project/vllm-ascend/pull/1800) |
| 16 | +* Prefix cache feature now work with the Ascend Scheduler. [#1446](https://github.com/vllm-project/vllm-ascend/pull/1446) |
| 17 | +* Torchair graph mode works with tp > 4 now. [#1508](https://github.com/vllm-project/vllm-ascend/issues/1508) |
| 18 | +* MTP support torchair graph mode now [#2145](https://github.com/vllm-project/vllm-ascend/pull/2145) |
| 19 | + |
| 20 | +## Other |
| 21 | + |
| 22 | +* Bug fixes: |
| 23 | + * Fix functional problem of multi-modality models like Qwen2-audio with Aclgraph. [#1803](https://github.com/vllm-project/vllm-ascend/pull/1803) |
| 24 | + * Fix the process group creating error with external launch scenario. [#1681](https://github.com/vllm-project/vllm-ascend/pull/1681) |
| 25 | + * Fix the functional problem with guided decoding. [#2022](https://github.com/vllm-project/vllm-ascend/pull/2022) |
| 26 | + * Fix the accuracy issue with common MoE models in DP scenario. [#1856](https://github.com/vllm-project/vllm-ascend/pull/1856) |
| 27 | +* Performance improved through a lot of prs: |
| 28 | + * Caching sin/cos instead of calculate it every layer. [#1890](https://github.com/vllm-project/vllm-ascend/pull/1890) |
| 29 | + * Improve shared expert multi-stream parallelism [#1891](https://github.com/vllm-project/vllm-ascend/pull/1891) |
| 30 | + * Implement the fusion of allreduce and matmul in prefill phase when tp is enabled. Enable this feature by setting `VLLM_ASCEND_ENABLE_MATMUL_ALLREDUCE` to `1`. [#1926](https://github.com/vllm-project/vllm-ascend/pull/1926) |
| 31 | + * Optimize Quantized MoE Performance by Reducing All2All Communication. [#2195](https://github.com/vllm-project/vllm-ascend/pull/2195) |
| 32 | + * Use AddRmsNormQuant ops in the custom model to optimize Qwen3's performance [#1806](https://github.com/vllm-project/vllm-ascend/pull/1806) |
| 33 | + * Use multicast to avoid padding decode request to prefill size [#1555](https://github.com/vllm-project/vllm-ascend/pull/1555) |
| 34 | + * The performance of LoRA has been improved. [#1884](https://github.com/vllm-project/vllm-ascend/pull/1884) |
| 35 | +* A batch of refactoring prs to enhance the code architecture: |
| 36 | + * Torchair model runner refactor [#2205](https://github.com/vllm-project/vllm-ascend/pull/2205) |
| 37 | + * Refactoring forward_context and model_runner_v1. [#1979](https://github.com/vllm-project/vllm-ascend/pull/1979) |
| 38 | + * Refactor AscendMetaData Comments. [#1967](https://github.com/vllm-project/vllm-ascend/pull/1967) |
| 39 | + * Refactor torchair utils. [#1892](https://github.com/vllm-project/vllm-ascend/pull/1892) |
| 40 | + * Refactor torchair worker. [#1885](https://github.com/vllm-project/vllm-ascend/pull/1885) |
| 41 | + * Register activation customop instead of overwrite forward_oot. [#1841](https://github.com/vllm-project/vllm-ascend/pull/1841) |
| 42 | +* Parameters changes: |
| 43 | + * `expert_tensor_parallel_size` in `additional_config` is removed now, and the EP and TP is aligned with vLLM now. [#1681](https://github.com/vllm-project/vllm-ascend/pull/1681) |
| 44 | + * Add `VLLM_ASCEND_MLA_PA` in environ variables, use this to enable mla paged attention operator for deepseek mla decode. |
| 45 | + * Add `VLLM_ASCEND_ENABLE_MATMUL_ALLREDUCE` in environ variables, enable `MatmulAllReduce` fusion kernel when tensor parallel is enabled. This feature is supported in A2, and eager mode will get better performance. |
| 46 | + * Add `VLLM_ASCEND_ENABLE_MOE_ALL2ALL_SEQ` in environ variables, Whether to enable moe all2all seq, this provides a basic framework on the basis of alltoall for easy expansion. |
| 47 | + |
| 48 | +* UT coverage reached 76.34% after a batch of prs followed by this rfc: [#1298](https://github.com/vllm-project/vllm-ascend/issues/1298) |
| 49 | +* Sequence Parallelism works for Qwen3 MoE. [#2209](https://github.com/vllm-project/vllm-ascend/issues/2209) |
| 50 | +* Chinese online document is added now. [#1870](https://github.com/vllm-project/vllm-ascend/issues/1870) |
| 51 | + |
| 52 | +### Known Issues |
| 53 | +* Aclgraph could not work with DP + EP currently, the mainly gap is the number of npu stream that Aclgraph needed to capture graph is not enough. [#2229](https://github.com/vllm-project/vllm-ascend/issues/2229) |
| 54 | +* There is an accuracy issue on W8A8 dynamic quantized DeepSeek with multistream enabled. This will be fixed in the next release. [#2232](https://github.com/vllm-project/vllm-ascend/issues/2232) |
| 55 | +* In Qwen3 MoE, SP cannot be incorporated into the Aclgraph. [#2246](https://github.com/vllm-project/vllm-ascend/issues/2246) |
| 56 | +* MTP not support V1 scheduler currently, will fix it in Q3. [#2254](https://github.com/vllm-project/vllm-ascend/issues/2254) |
| 57 | +* When running MTP with DP > 1, we need to disable metrics logger due to some issue on vLLM. [#2254](https://github.com/vllm-project/vllm-ascend/issues/2254) |
| 58 | + |
3 | 59 | ## v0.9.1rc2 - 2025.08.04 |
4 | 60 | This is the 2nd release candidate of v0.9.1 for vLLM Ascend. Please follow the [official doc](https://vllm-ascend.readthedocs.io/en/v0.9.1-dev/) to get started. |
5 | 61 |
|
|
0 commit comments