A study of manycore shared memory architecture as a way to build SOC applications

Yosi Ben Asher, Yousef Shajrawi, Y. Gendel, Gadi Haber, Oren Segal

Research output: Contribution to journalConference articlepeer-review

Abstract

Manycore shared memory architectures hold a significant premise to speed up and simplify SOCs. Using many homogeneous small-cores will allow replacing the hardware accelerators of SOCs by parallel algorithms communicating through shared memory. Currently shared memory is realized by maintaining cache-consistency across the cores, caching all the connected cores to one main memory module. This approach, though used today, is not likely to be scalable enough to support the high number of cores needed for highly parallel SOCs. Therefore we consider a theoretical scheme for shared memory wherein: the shared address space is divided between a set of memory modules; and a communication network allows each core to access every such module in parallel. Load-balancing between the memory modules is obtained by rehashing the memory address-space. We consider practical aspects involved with a practical realization of this scheme, e.g. how will the wire complexity of the communication network affect the execution time. We have designed a simple generic shared memory architecture, synthesized it to 2,4,8,16,., 1024-cores for FPGA virtex-7 and evaluated it on several parallel programs. The synthesis results and the execution measurements show that, for the FPGA. all problematic aspects of this construction can be resolved. For example, unlike ASICs, the growing complexity of the communication network is absorbed by the FPGA's routing grid and by its routing mechanism. This makes this type of architectures particularly suitable for FPGAs. We used 32-bits modified PACOBLAZE cores and tested different parameters of this architecture verifying its ability to achieve high speedups. The results suggest that re-hashing is not essential and one hash-function suffice (compared to the family of universal hash functions that is needed by the theoretical construction).

Original languageAmerican English
Pages (from-to)174-181
Number of pages8
JournalSimulation Series
Volume47
Issue number4
StatePublished - 2015
Event23rd High Performance Computing Symposium, HPC 2015, Part of the 2015 Spring Simulation Multi-Conference, SpringSim 2015 - Alexandria, United States
Duration: 12 Apr 201515 Apr 2015

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'A study of manycore shared memory architecture as a way to build SOC applications'. Together they form a unique fingerprint.

Cite this