arXiv cs.AI· Prabod Rathnayaka, Isura Manchanayake, Srikandabala Kogul, Steve Siu, Fabian Waschkowski, Lukas Wesemann·· 4 小时前AI 评分59
Superfluid 论文发布:面向本地分布式 LLM 推理的开源服务器
Superfluid: Local AI Server for Distributed LLM Inference
AI 导读
研究团队发布开源本地 LLM 推理服务器 Superfluid(arXiv:2611.17753),在单机上通过统一调度器优先处理交互请求、支持多智能体共享本地模型,并可把多台不同厂商的机器整合为单一 API 端点。
正文 · 原文
Abstract:Running large language models (LLMs) on local devices keeps data private, removes per-token cost, cuts network latency, and keeps working without a connection. Two shifts now strain how such models are served. On one device, several agents share a locally served model with an interactive user. Across devices, a developer or small team often owns several capable machines from different vendors, which sit idle most of the day because every local server serves one runtime on one machine. We present Superfluid, an open-source LLM server built for local inference at both scales. On one device, Superfluid drives this http URL, MLX, or BaseRT in isolated worker processes through one scheduler that moves interactive requests ahead of agent work, computes a prompt shared by many requests once, and writes every token to a durable log before it reaches the client. Across devices, one Superfluid server turns a fleet of machines into a single API endpoint. Machines with different hardware serve copies of the same model, receive work according to their measured speed, and can join or fail under load without losing a request. For Qwen3.8-27B on an Apple M5 Pro, a chat sent while eight agents are working receives its first token in 1.5 s, against 291-612 s on llama-server, Ollama, and this http URL. On six machines from three chip vendors serving 64 simulated users for one hour, Superfluid answers all 5,041 requests while machines join one at a time and one is killed and rejoins; throughput grows from 112 to 534 tok/s and the median first-token time falls from 99 to 12 s. On twelve Mac Studio M3 Ultras it sustains 10,656 tok/s across 38,563 requests from 384 agents without a failure. Superfluid is available under the Apache-2.0 licence at this https URL.
| Subjects: | Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.17753 [cs.DC] |
| (or arXiv:2611.17753v1 [cs.DC] for this version) | |
| https://doi.org/10.48550/arXiv.2611.17753 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Prabod Rathnayaka [view email]
[v1]
Fri, 13 Nov 2026 05:10:32 UTC (89 KB)
来源:arXiv cs.AI · arxiv.org