Benchmarking Memory-Centric Computing Systems: Analysis of Real Processing-in-Memory Hardware

TL;DR

This paper analyzes the first real-world processing-in-memory (PIM) architecture by UPMEM, highlighting its performance, workload suitability, and design insights for future PIM systems, addressing the memory bottleneck in modern workloads.

Contribution

It provides the first comprehensive analysis of UPMEM's real-world PIM architecture, offering practical insights and recommendations for hardware and software design.

Findings

UPMEM PIM architecture effectively reduces data movement overhead.

Workload characteristics influence PIM system performance.

Guidelines for optimizing software and hardware for PIM systems.

Abstract

Many modern workloads such as neural network inference and graph processing are fundamentally memory-bound. For such workloads, data movement between memory and CPU cores imposes a significant overhead in terms of both latency and energy. A major reason is that this communication happens through a narrow bus with high latency and limited bandwidth, and the low data reuse in memory-bound workloads is insufficient to amortize the cost of memory access. Fundamentally addressing this data movement bottleneck requires a paradigm where the memory system assumes an active role in computing by integrating processing capabilities. This paradigm is known as processing-in-memory (PIM). Recent research explores different forms of PIM architectures, motivated by the emergence of new technologies that integrate memory with a logic layer, where processing elements can be easily placed. Past works evaluate these architectures in simulation or, at best, with simplified hardware prototypes. In contrast, the UPMEM company has designed and manufactured the first publicly-available real-world PIM architecture. The UPMEM PIM architecture combines traditional DRAM memory arrays with general-purpose in-order cores, called DRAM Processing Units (DPUs), integrated in the same chip. This paper presents key takeaways from the first comprehensive analysis of the first publicly-available real-world PIM architecture. We provide four key takeaways about the UPMEM PIM architecture, which stem from our study. More insights about suitability of different workloads to the PIM system, programming recommendations for software designers, and suggestions and hints for hardware and architecture designers of future PIM systems are available in arXiv:2105.03814