Salvato in:
Dettagli Bibliografici
Autori principali: Barreto, João, Castro, Daniel, Romano, Paolo, Baldassin, Alexandro
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2410.16110
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914982152634368
author Barreto, João
Castro, Daniel
Romano, Paolo
Baldassin, Alexandro
author_facet Barreto, João
Castro, Daniel
Romano, Paolo
Baldassin, Alexandro
contents Despite the recent improvements in supporting Persistent Hardware Transactions (PHTs) on emerging persistent memories (PM), the poor performance of Read-Only (RO) transactions remains largely overlooked. We propose DUMBO, a new design for PHT that eliminates the two most crucial bottlenecks that hinder RO transactions in state-of-the-art PHT. At its core, DUMBO exploits advanced instructions that some contemporary HTMs provide to suspend (and resume) transactional access tracking. Our experimental evaluation with an IBM POWER9 system using the TPC-C benchmark shows that DUMBO can outperform the state of the art designs for persistent hardware (SPHT) and software memory transactions (Pisces), by up to 4.0x.
format Preprint
id arxiv_https___arxiv_org_abs_2410_16110
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DUMBO: Making durable read-only transactions fly on hardware transactional memory
Barreto, João
Castro, Daniel
Romano, Paolo
Baldassin, Alexandro
Distributed, Parallel, and Cluster Computing
Despite the recent improvements in supporting Persistent Hardware Transactions (PHTs) on emerging persistent memories (PM), the poor performance of Read-Only (RO) transactions remains largely overlooked. We propose DUMBO, a new design for PHT that eliminates the two most crucial bottlenecks that hinder RO transactions in state-of-the-art PHT. At its core, DUMBO exploits advanced instructions that some contemporary HTMs provide to suspend (and resume) transactional access tracking. Our experimental evaluation with an IBM POWER9 system using the TPC-C benchmark shows that DUMBO can outperform the state of the art designs for persistent hardware (SPHT) and software memory transactions (Pisces), by up to 4.0x.
title DUMBO: Making durable read-only transactions fly on hardware transactional memory
topic Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2410.16110