Quantitative comparison of hardware transactional memory for Blue Gene/Q, zEnterprise EC12, Intel Core, and POWER8
Takuya Nakaike, Rei Odaira, Matthew Gaudet, Maged M. Michael, Hisanobu Tomari
Abstract
Takuya Nakaike, Rei Odaira, Matthew Gaudet, Maged M. Michael, Hisanobu Tomari
Abstract
Transactional Memory (TM) is a new programming paradigm for both simple concurrent programming and high concurrent performance. Hardware Transactional Memory (HTM) is hardware support for TM-based programming. It has lower overhead than software transactional memory (STM), which is a software-based implementation of TM. There are now four commercial systems, IBM Blue Gene/Q, IBM zEnterprise EC12, Intel Core, and IBM POWER8, offering HTM. Our work is the first to compare the performance of these four HTM systems. We measured the STAMP benchmarks, the most widely used TM benchmarks. We also evaluated the specific features of each HTM system. Our experimental results show that: (1) there is no single HTM system that is more scalable than the others in all of the benchmarks, (2) there are measurable performance differences among the HTM systems in some benchmarks, and (3) each HTM system has its own implementation characteristics that limit its scalability.
OpenAlex reports 83 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Transactional Memory (TM) is a new programming paradigm for both simple concurrent programming and high concurrent performance. Hardware Transactional Memory (HTM) is hardware support for TM-based programming. It has lower overhead than software transactional memory (STM), which is a software-based implementation of TM. There are now four commercial systems, IBM Blue Gene/Q, IBM zEnterprise EC12, Intel Core, and IBM POWER8, offering HTM. Our work is the first to compare the performance of these four HTM systems. We measured the STAMP benchmarks, the most widely used TM benchmarks. We also evaluated the specific features of each HTM system. Our experimental results show that: (1) there is no single HTM system that is more scalable than the others in all of the benchmarks, (2) there are measurable performance differences among the HTM systems in some benchmarks, and (3) each HTM system has its own implementation characteristics that limit its scalability.
Key concepts: Transactional memory, Computer science, IBM, Software transactional memory, Scalability, Operating system, Overhead (engineering), Software