H
Harold W. Cain
Researcher at IBM
Publications - 112
Citations - 2409
Harold W. Cain is an academic researcher from IBM. The author has contributed to research in topics: Cache & Transactional memory. The author has an hindex of 26, co-authored 112 publications receiving 2386 citations. Previous affiliations of Harold W. Cain include University of Wisconsin-Madison & Qualcomm.
Papers
More filters
Journal ArticleDOI
Software transactional memory: why is it only a research toy?
Calin Cascaval,Colin Blundell,Maged M. Michael,Harold W. Cain,Peng Wu,Stefanie R. Chiras,Siddhartha Chatterjee +6 more
TL;DR: The promise of STM may likely be undermined by its overheads and workload applicabilities.
Journal ArticleDOI
Software Transactional Memory: Why Is It Only a Research Toy?: The promise of STM may likely be undermined by its overheads and workload applicabilities.
Calin Cascaval,Colin Blundell,Maged M. Michael,Harold W. Cain,Peng Wu,Stefanie R. Chiras,Siddhartha Chatterjee +6 more
TL;DR: It is observed that the overall performance of TM is significantly worse at low levels of parallelism, which is likely to limit the adoption of this programming paradigm.
Proceedings ArticleDOI
An architectural evaluation of Java TPC-W
TL;DR: This work presents a detailed characterization of the Transaction Processing Council's TPC-W web benchmark, implemented in Java, and presents an architectural study detailing the memory system and branch predictor behavior of the workload.
Proceedings ArticleDOI
Robust architectural support for transactional memory in the power architecture
TL;DR: Architectural support for transactional memory added to a future version of the Power ISA™ is described, including some previously unexplored interactions between TM and existing features of the ISA, and the motivation and rationale for the choices of architectural semantics are discussed.
Journal ArticleDOI
Accurate, efficient, and adaptive calling context profiling
TL;DR: This work develops a new approach for building the calling context tree at runtime, called adaptive bursting, which dramatically reduces overhead while preserving profile accuracy and achieves 85% degree of overlap.