On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable Processors

Reconfigurable instruction set processors provide the possibility of tailor the instruction set of a CPU to a particular application. While this customization process could be performed during runtime in order to adapt the CPU to the currently executed workload, this use case has been hardly investi...

Full description

Bibliographic Details
Main Authors: Mariusz Grad, Christian Plessl
Format: Article
Language:English
Published: Hindawi Limited 2012-01-01
Series:International Journal of Reconfigurable Computing
Online Access:http://dx.doi.org/10.1155/2012/418315
id doaj-01577768fc234d85bd85e93d0b8ccb4d
record_format Article
spelling doaj-01577768fc234d85bd85e93d0b8ccb4d2020-11-25T01:40:08ZengHindawi LimitedInternational Journal of Reconfigurable Computing1687-71951687-72092012-01-01201210.1155/2012/418315418315On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable ProcessorsMariusz Grad0Christian Plessl1Paderborn Center for Parallel Computing, University of Paderborn, 33098 Paderborn, GermanyPaderborn Center for Parallel Computing, University of Paderborn, 33098 Paderborn, GermanyReconfigurable instruction set processors provide the possibility of tailor the instruction set of a CPU to a particular application. While this customization process could be performed during runtime in order to adapt the CPU to the currently executed workload, this use case has been hardly investigated. In this paper, we study the feasibility of moving the customization process to runtime and evaluate the relation of the expected speedups and the associated overheads. To this end, we present a tool flow that is tailored to the requirements of this just-in-time ASIP specialization scenario. We evaluate our methods by targeting our previously introduced Woolcano reconfigurable ASIP architecture for a set of applications from the SPEC2006, SPEC2000, MiBench, and SciMark2 benchmark suites. Our results show that just-in-time ASIP specialization is promising for embedded computing applications, where average speedups of 5x can be achieved by spending 50 minutes for custom instruction identification and hardware generation. These overheads will be compensated if the applications execute for more than 2 hours. For the scientific computing benchmarks, the achievable speedup is only 1.2x, which requires significant execution times in the order of days to amortize the overheads.http://dx.doi.org/10.1155/2012/418315
collection DOAJ
language English
format Article
sources DOAJ
author Mariusz Grad
Christian Plessl
spellingShingle Mariusz Grad
Christian Plessl
On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable Processors
International Journal of Reconfigurable Computing
author_facet Mariusz Grad
Christian Plessl
author_sort Mariusz Grad
title On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable Processors
title_short On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable Processors
title_full On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable Processors
title_fullStr On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable Processors
title_full_unstemmed On the Feasibility and Limitations of Just-in-Time Instruction Set Extension for FPGA-Based Reconfigurable Processors
title_sort on the feasibility and limitations of just-in-time instruction set extension for fpga-based reconfigurable processors
publisher Hindawi Limited
series International Journal of Reconfigurable Computing
issn 1687-7195
1687-7209
publishDate 2012-01-01
description Reconfigurable instruction set processors provide the possibility of tailor the instruction set of a CPU to a particular application. While this customization process could be performed during runtime in order to adapt the CPU to the currently executed workload, this use case has been hardly investigated. In this paper, we study the feasibility of moving the customization process to runtime and evaluate the relation of the expected speedups and the associated overheads. To this end, we present a tool flow that is tailored to the requirements of this just-in-time ASIP specialization scenario. We evaluate our methods by targeting our previously introduced Woolcano reconfigurable ASIP architecture for a set of applications from the SPEC2006, SPEC2000, MiBench, and SciMark2 benchmark suites. Our results show that just-in-time ASIP specialization is promising for embedded computing applications, where average speedups of 5x can be achieved by spending 50 minutes for custom instruction identification and hardware generation. These overheads will be compensated if the applications execute for more than 2 hours. For the scientific computing benchmarks, the achievable speedup is only 1.2x, which requires significant execution times in the order of days to amortize the overheads.
url http://dx.doi.org/10.1155/2012/418315
work_keys_str_mv AT mariuszgrad onthefeasibilityandlimitationsofjustintimeinstructionsetextensionforfpgabasedreconfigurableprocessors
AT christianplessl onthefeasibilityandlimitationsofjustintimeinstructionsetextensionforfpgabasedreconfigurableprocessors
_version_ 1725046912901447680