34 min listen
PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU
PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU
ratings:
Length:
52 minutes
Released:
Dec 28, 2023
Format:
Podcast episode
Description
This paper introduces PowerInfer, a high-speed Large Language Model (LLM) inference engine on a personal computer (PC) equipped with a single consumer-grade GPU. The key underlying the design of PowerInfer is exploiting the high locality inherent in LLM inference, characterized by a power-law distribution in neuron activation. This distribution indicates that a small subset of neurons, termed hot neurons, are consistently activated across inputs, while the majority, cold neurons, vary based on specific inputs. PowerInfer exploits such an insight to design a GPU-CPU hybrid inference engine: hot-activated neurons are preloaded onto the GPU for fast access, while cold-activated neurons are computed on the CPU, thus significantly reducing GPU memory demands and CPU-GPU data transfers. PowerInfer further integrates adaptive predictors and neuron-aware sparse operators, optimizing the efficiency of neuron activation and computational sparsity. Evaluation shows that PowerInfer attains an average token generation rate of 13.20 tokens/s, with a peak of 29.08 tokens/s, across various LLMs (including OPT-175B) on a single NVIDIA RTX 4090 GPU, only 18% lower than that achieved by a top-tier server-grade A100 GPU. This significantly outperforms llama.cpp by up to 11.69x while retaining model accuracy.
2023: Yixin Song, Zeyu Mi, Haotong Xie, Haibo Chen
https://arxiv.org/pdf/2312.12456v1.pdf
2023: Yixin Song, Zeyu Mi, Haotong Xie, Haibo Chen
https://arxiv.org/pdf/2312.12456v1.pdf
Released:
Dec 28, 2023
Format:
Podcast episode
Titles in the series (100)
Editing Large Language Models: Problems, Methods, and Opportunities: Recent advancements in deep learning have precipitated the emergence of large language models (LLMs) which exhibit an impressive aptitude for understanding and producing text akin to human language. Despite the ability to train highly capable LLMs, t... by Papers Read on AI