UpDown: Programmable fine-grained Events for Scalable Performance on Irregular Applications

TL;DR

UpDown is a novel accelerator architecture designed to efficiently handle irregular applications with fine-grained data and control flows, achieving significant performance improvements over CPUs and prior accelerators.

Contribution

It introduces hardware primitives like lightweight threading and event-driven scheduling for programmable fine-grained execution in irregular applications.

Findings

UpDown outperforms CPUs by up to 195x in graph analytics.

It achieves over 4x speedup compared to prior accelerators.

Generates high memory parallelism (~4.6x over CPU).

Abstract

Applications with irregular data structures, data-dependent control flows and fine-grained data transfers (e.g., real-world graph computations) perform poorly on cache-based systems. We propose the UpDown accelerator that supports fine-grained execution with novel architecture mechanisms - lightweight threading, event-driven scheduling, efficient ultra-short threads, and split-transaction DRAM access with software-controlled synchronization. These hardware primitives support software programmable events, enabling high performance on diverse data structures and algorithms. UpDown also supports scalable performance; hardware replication enables programs to scale up performance. Evaluation results show UpDown's flexibility and scalability enable it to outperform CPUs on graph mining and analytics computations by up to 116-195x geomean speedup and more than 4x speedup over prior accelerators. We show that UpDown generates high memory parallelism (~4.6x over CPU) required for memory intensive graph computations. We present measurements that attribute the performance of UpDown (23x architectural advantage) to its individual architectural mechanisms. Finally, we also analyze the area and power cost of UpDown's mechanisms for software programmability.