Cloud Server Benchmarks for Performance Evaluation of New Hardware Architecture

TL;DR

This paper introduces PALMScloud, a new suite of cloud benchmarks designed for performance evaluation of cloud servers and compatible with gem5 simulation, addressing limitations of existing benchmarks in representing typical cloud workloads.

Contribution

The paper presents PALMScloud, a novel cloud benchmark suite that accurately models cloud workloads and is compatible with cycle-accurate simulators like gem5, enabling better hardware evaluation.

Findings

PALMScloud effectively evaluates cache performance in cloud servers.

The suite is compatible with gem5 and real machines.

Case study demonstrates its use in assessing a new secure cache.

Abstract

Adding new hardware features to a cloud computing server requires testing both the functionalities and the performance of the new hardware mechanisms. However, commonly used cloud computing server workloads are not well-represented by the SPEC integer and floating-point benchmark and Parsec suites typically used by the computer architecture community. Existing cloud benchmark suites for scale-out or scale-up computing are not representative of the most common cloud usage, and are very difficult to run on a cycle-accurate simulator that can accurately model new hardware, like gem5. In this paper, we present PALMScloud, a suite of cloud computing benchmarks for performance evaluation of cloud servers, that is ready to run on the gem5 cycle-accurate simulator. We demonstrate how our cloud computing benchmarks are used in evaluating the cache performance of a new secure cache called Newcache as a case study. We hope that these cloud benchmarks, ready to run on a dual-machine gem5 simulator or on real machines, can be useful to other researchers interested in improving hardware micro-architecture and cloud server performance.