Serving frontier MoE and multimodal models on Ascend 910 via vLLM-Ascend is feasible but dominated by engineering cost from incomplete operators, fragile parallelism, kernel faults, and weak observability.
vLLM Ascend plugin (vllm-ascend) documentation.https://vllm-ascend
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
On the Limitations of Non-GPU AI Accelerators for Large-Model Inference: A Field Study of MoE and Multimodal Serving on Huawei Ascend
Serving frontier MoE and multimodal models on Ascend 910 via vLLM-Ascend is feasible but dominated by engineering cost from incomplete operators, fragile parallelism, kernel faults, and weak observability.