Presentation 2010-12-17
Fine Grained Time Sharing to Extend Capacity of FU Array
Takuya IWAKAMI, Kazuhiro YOSHIMURA, Kodai MORI, TAKASHI Nakada, Yasuhiko NAKASHIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We have previously proposed Linear Array Pipeline Processor (LAPP) which can map an inner loop of conventional VLIW codes onto Function Unit (FU) array and use minimum required FUs to exploit performance per watt. However, by this structure, LAPP has a constraint that the maximum codes to be mapped define the number of FUs. Though a large number of FUs can be predetermined, it will consequently prevent the power, area and performance balance. To solve this problem, we propose a fine-grained time sharing mechanism to extend the FU array of LAPP, to remove from the physical constraint due to the limited size of FU array. We evaluate the perfor-mance, power consumption and area of the extended LAPP. The results indicate that the extended LAPP executes about 6-fold faster than the baseline LAPP. The performance per area is increased by 16 % and the performance per watt is reduced by 20 % accordingly.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) time-sharing / FU array / reconfigurable architecture
Paper # ICD2010-123
Date of Issue

Conference Information
Committee ICD
Conference Date 2010/12/9(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Integrated Circuits and Devices (ICD)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Fine Grained Time Sharing to Extend Capacity of FU Array
Sub Title (in English)
Keyword(1) time-sharing
Keyword(2) FU array
Keyword(3) reconfigurable architecture
1st Author's Name Takuya IWAKAMI
1st Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology()
2nd Author's Name Kazuhiro YOSHIMURA
2nd Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
3rd Author's Name Kodai MORI
3rd Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
4th Author's Name TAKASHI Nakada
4th Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
5th Author's Name Yasuhiko NAKASHIMA
5th Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
Date 2010-12-17
Paper # ICD2010-123
Volume (vol) vol.110
Number (no) 344
Page pp.pp.-
#Pages 6
Date of Issue