Predictable LLM Serving on GPU Clusters
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914010704642048 |
|---|---|
| author | Darzi, Erfan Bharadwaj, Shreeanant Balija, Sree Bhargavi |
| author_facet | Darzi, Erfan Bharadwaj, Shreeanant Balija, Sree Bhargavi |
| contents | Latency-sensitive inference on shared A100 clusters often suffers noisy-neighbor interference on the PCIe fabric, inflating tail latency and SLO violations. We present a fabric-agnostic, VM-deployable host-level controller that combines dynamic Multi-Instance GPU (MIG) reconfiguration, PCIe-aware placement, and lightweight guardrails (MPS quotas, cgroup I/O). It samples per-tenant tails and system signals, uses topology hints to avoid PCIe hot spots, and gates actions with dwell/cool-down to avoid thrash. On a single host and a 2-node (16-GPU) cluster, SLO miss-rate is reduced by \(\approx\)32\% (\(\approx\)1.5) and p99 latency improves \(\approx\)15\% with \(\leq\)5\% throughput cost versus static MIG and naive placement; ablations show MIG and placement contribute comparably. We also evaluate LLM serving with vLLM on OLMo 2 7B Instruct: TTFT p99 improves \(\approx\)10--15\% at \(\leq\)5\% cost without changing the controller. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_20274 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Predictable LLM Serving on GPU Clusters Darzi, Erfan Bharadwaj, Shreeanant Balija, Sree Bhargavi Distributed, Parallel, and Cluster Computing Latency-sensitive inference on shared A100 clusters often suffers noisy-neighbor interference on the PCIe fabric, inflating tail latency and SLO violations. We present a fabric-agnostic, VM-deployable host-level controller that combines dynamic Multi-Instance GPU (MIG) reconfiguration, PCIe-aware placement, and lightweight guardrails (MPS quotas, cgroup I/O). It samples per-tenant tails and system signals, uses topology hints to avoid PCIe hot spots, and gates actions with dwell/cool-down to avoid thrash. On a single host and a 2-node (16-GPU) cluster, SLO miss-rate is reduced by \(\approx\)32\% (\(\approx\)1.5) and p99 latency improves \(\approx\)15\% with \(\leq\)5\% throughput cost versus static MIG and naive placement; ablations show MIG and placement contribute comparably. We also evaluate LLM serving with vLLM on OLMo 2 7B Instruct: TTFT p99 improves \(\approx\)10--15\% at \(\leq\)5\% cost without changing the controller. |
| title | Predictable LLM Serving on GPU Clusters |
| topic | Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2508.20274 |