Single-graph multiple flows: Energy efficient design alternative for GPGPUs
Authors:
- 1 June 2014
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 10636897,p. 205-216
Abstract
We present the single-graph multiple-flows (SGMF) architecture that combines coarse-grain reconfigurable computing with dynamic dataflow to deliver massive thread-level parallelism. The CUDA-compatible SGMF architecture is positioned as an energy efficient design alternative for GPGPUs. The architecture maps a compute kernel, represented as a dataflow graph, onto a coarse-grain reconfigurable fabric composed of a grid of interconnected functional units. Each unit dynamically schedules instances of the same static instruction originating from different CUDA threads. The dynamically scheduled functional units enable streaming the data of multiple threads (or graph flows, in SGMF parlance) through the grid. The combination of statically mapped instructions and direct communication between functional units obviate the need for afull instruction pipeline and a centralized register file, whose energy overheads burden GPGPUs. We show that the SGMF architecture delivers performance comparable to that of contemporary GPGPUs while consuming ~57% less energy on average.Keywords
This publication has 37 references indexed in Scilit:
- Scalable Parallel Programming with CUDAQueue, 2008
- The Raw microprocessor: a computational fabric for software circuits and general-purpose programsIEEE Micro, 2002
- Imagine: media processing with streamsIEEE Micro, 2001
- PipeRench: a reconfigurable architecture and compilerComputer, 2000
- ARB: a hardware mechanism for dynamic reordering of memory referencesIEEE Transactions on Computers, 1996
- Efficiently computing static single assignment form and the control dependence graphACM Transactions on Programming Languages and Systems, 1991
- Properties and performance of folded hypercubesIEEE Transactions on Parallel and Distributed Systems, 1991
- Executing a program on the MIT tagged-token dataflow architectureIEEE Transactions on Computers, 1990
- The Manchester prototype dataflow computerCommunications of the ACM, 1985
- The CRAY-1 computer systemCommunications of the ACM, 1978