Near-Memory Address Translation

TL;DR

This paper introduces DIPTA, a near-memory address translation structure that enables parallel translation and data fetch, significantly improving performance for memory-side processing units by eliminating translation overhead.

Contribution

It proposes DIPTA, a novel near-memory translation structure that simplifies mapping and enables parallel translation and data fetch, overcoming limitations of traditional TLBs.

Findings

DIPTA achieves up to 3.81x speedup over conventional translation.

Limiting virtual-to-physical mapping associativity incurs no penalty.

Parallel translation and data fetch improve MPU applicability in memory systems.

Abstract

Memory and logic integration on the same chip is becoming increasingly cost effective, creating the opportunity to offload data-intensive functionality to processing units placed inside memory chips. The introduction of memory-side processing units (MPUs) into conventional systems faces virtual memory as the first big showstopper: without efficient hardware support for address translation MPUs have highly limited applicability. Unfortunately, conventional translation mechanisms fall short of providing fast translations as contemporary memories exceed the reach of TLBs, making expensive page walks common. In this paper, we are the first to show that the historically important flexibility to map any virtual page to any page frame is unnecessary in today's servers. We find that while limiting the associativity of the virtual-to-physical mapping incurs no penalty, it can break the translate-then-fetch serialization if combined with careful data placement in the MPU's memory, allowing for translation and data fetch to proceed independently and in parallel. We propose the Distributed Inverted Page Table (DIPTA), a near-memory structure in which the smallest memory partition keeps the translation information for its data share, ensuring that the translation completes together with the data fetch. DIPTA completely eliminates the performance overhead of translation, achieving speedups of up to 3.81x and 2.13x over conventional translation using 4KB and 1GB pages respectively.