Automatic Data Partitioning in Software Transactional Memories
Research output: Contribution to conferences › Paper › Contributed › peer-review
Contributors
Abstract
We investigate to which extent data partitioning can help
improve the performance of software transactional memory
(STM). Our main idea is that the access patterns of the var-
ious data structures of an application might be sufficiently
different so that it would be beneficial to tune the behav-
ior of the STM for individual data partitions. We evaluate
our approach using standard transactional memory bench-
marks. We show that these applications contain partitions
with different characteristics and, despite the runtime over-
head introduced by partition tracking and dynamic tuning,
that partitioning provides significant performance improve-
ments.
improve the performance of software transactional memory
(STM). Our main idea is that the access patterns of the var-
ious data structures of an application might be sufficiently
different so that it would be beneficial to tune the behav-
ior of the STM for individual data partitions. We evaluate
our approach using standard transactional memory bench-
marks. We show that these applications contain partitions
with different characteristics and, despite the runtime over-
head introduced by partition tracking and dynamic tuning,
that partitioning provides significant performance improve-
ments.
Details
Original language | English |
---|---|
Pages | 152-159 |
Number of pages | 8 |
Publication status | Published - 2008 |
Peer-reviewed | Yes |
Conference
Title | SPAA '08: twentieth annual symposium on Parallelism in algorithms and architectures, ACM, 2008 |
---|---|
Abbreviated title | SPAA '08 |
Conference number | |
Duration | 14 June 2008 |
Degree of recognition | International event |
Location | |
City | München |
Country | Germany |
External IDs
Scopus | 57349083101 |
---|