Litcius/Paper detail

NeuPIMs: NPU-PIM Heterogeneous Acceleration for Batched LLM Inferencing

Guseul Heo, Sangyeop Lee, Jaehong Cho, Hyunmin Choi, S. K. Lee, Hyungkyu Ham, Gwangsun Kim, Divya Mahajan, Jongse Park

202492 citationsDOIOpen Access PDF

Abstract

Modern transformer-based Large Language Models (LLMs) are constructed with a series of decoder blocks. Each block comprises three key components: (1) QKV generation, (2) multi-head attention, and (3) feed-forward networks. In batched processing, QKV generation and feed-forward networks involve compute-intensive matrix-matrix multiplications (GEMM), while multi-head attention requires bandwidth-heavy matrix-vector multiplications (GEMV). Machine learning accelerators like TPUs or NPUs are proficient in handling GEMM but are less efficient for GEMV computations. Conversely, Processing-in-Memory (PIM) technology is tailored for efficient GEMV computation, while it lacks the computational power to handle GEMM effectively.

Topics & Concepts

Computer scienceParallel computingComputationInterleavingBlock (permutation group theory)Key (lock)Distributed computingAlgorithmOperating systemGeometryMathematicsFerroelectric and Negative Capacitance DevicesTopic ModelingAdvanced Data Storage Technologies