SpecEdge: Scalable Edge-Assisted Serving Framework for Interactive LLMs
원문 초록
Large language models (LLMs) power many modern applications, but serving them at scale remains costly and resource-intensive. Current server-centric systems overlook consumer-grade GPUs at the edge. We introduce SpecEdge, an edge-assisted inference framework that splits LLM workloads between edge and server GPUs using a speculative decoding scheme, exchanging only token outputs over the network. SpecEdge employs proactive edge drafting to overlap edge token creation with server verification and pipeline-aware scheduling that interleaves multiple user requests to increase server-side throughput. Experiments show SpecEdge enhances overall cost efficiency by 1.91x through achieving 2.22x server throughput, and reduces inter token latency by 11.24% compared to a server-only baseline, introducing a scalable, cost-effective paradigm for LLM serving.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Optimal Sensor Scheduling and Selection for Continuous-Discrete Kalman Filtering with Auxiliary Dynamics
SystolicAttention: Fusing FlashAttention within a Single Systolic Array
ANN-Based Grid Impedance Estimation for Adaptive Gain Scheduling in VSG Under Dynamic Grid Conditions
Confucius3-Math: A Lightweight High-Performance Reasoning LLM for Chinese K-12 Mathematics Learning
ConsumerBench: Benchmarking Generative AI Applications on End-User Devices
RapFlow-TTS: Rapid and High-Fidelity Text-to-Speech with Improved Consistency Flow Matching
토론
이 논문과 연결된 포럼 글이 아직 없습니다.