IANUS: Integrated Accelerator based on NPU-PIM Unified Memory System

TL;DR

IANUS is a novel integrated accelerator combining NPU and PIM with a unified memory system to efficiently accelerate end-to-end LLM inference, significantly outperforming existing GPU and accelerator solutions.

Contribution

The paper introduces IANUS, a domain-specific architecture that unifies NPU and PIM with a shared memory system and novel scheduling to enhance LLM inference performance.

Findings

IANUS improves GPT-2 inference speed by 6.2x over NVIDIA A100.

It achieves a 3.2x average speedup compared to the state-of-the-art accelerator.

Prototype implementation demonstrates feasibility with commercial PIM, NPU, and FPGA.

Abstract

Accelerating end-to-end inference of transformer-based large language models (LLMs) is a critical component of AI services in datacenters. However, diverse compute characteristics of end-to-end LLM inference present challenges as previously proposed accelerators only address certain operations or stages (e.g., self-attention, generation stage, etc.). To address the unique challenges of accelerating end-to-end inference, we propose IANUS -- Integrated Accelerator based on NPU-PIM Unified Memory System. IANUS is a domain-specific system architecture that combines a Neural Processing Unit (NPU) with a Processing-in-Memory (PIM) to leverage both the NPU's high computation throughput and the PIM's high effective memory bandwidth. In particular, IANUS employs a unified main memory system where the PIM memory is used both for PIM operations and for NPU's main memory. The unified main memory system ensures that memory capacity is efficiently utilized and the movement of shared data between NPU and PIM is minimized. However, it introduces new challenges since normal memory accesses and PIM computations cannot be performed simultaneously. Thus, we propose novel PIM Access Scheduling that manages normal memory accesses and PIM computations through workload mapping and scheduling across the PIM and the NPU. Our detailed simulation evaluations show that IANUS improves the performance of GPT-2 by 6.2$\times$ and 3.2$\times$, on average, compared to the NVIDIA A100 GPU and the state-of-the-art accelerator. As a proof-of-concept, we develop a prototype of IANUS with a commercial PIM, NPU, and an FPGA-based PIM controller to demonstrate the feasibility of IANUS.