Merging VLIW and vector processing techniques for a simple, high-performance processor architecture. Issue 7 (July 2015)
- Record Type:
- Journal Article
- Title:
- Merging VLIW and vector processing techniques for a simple, high-performance processor architecture. Issue 7 (July 2015)
- Main Title:
- Merging VLIW and vector processing techniques for a simple, high-performance processor architecture
- Authors:
- Soliman, Mostafa I.
- Abstract:
- <abstract abstract-type="author" id="ab0005"> <title id="sect0005">Abstract</title> <sec> <p id="sp0130">This paper proposes a new processor architecture called VVSHP for accelerating data-parallel applications, which are growing in importance and demanding increased performance from hardware. VVSHP merges VLIW and vector processing techniques for a simple, high-performance processor architecture. One key point of VVSHP is the execution of multiple scalar instructions within VLIW and vector instructions on unified parallel execution datapaths. Another key point is to reduce the complexity of VVSHP by designing a two-part register file: (1) shared scalar–vector part with eight-read/four-write ports 64×32-bit registers (64 scalar or 16×4 vector registers) for storing scalar/vector data and (2) vector part with two-read/one-write ports 48 vector-registers, each stores 4×32-bit vector data. Moreover, processing vector data with lengths varying from 1 to 256 represents a key point for reducing the loop overheads. VVSHP can issue up to four scalar/vector operations in each cycle for parallel processing a set of operands and producing up to four results to be written back into VVSHP register file. However, it cannot issue more than one memory operation at a time, which loads/stores 128-bit scalar/vector data from/to data memory. The design of our proposed VVSHP processor is implemented using VHDL targeting the Xilinx FPGA Virtex-5 and its performance is evaluated.</p> </sec><abstract abstract-type="author" id="ab0005"> <title id="sect0005">Abstract</title> <sec> <p id="sp0130">This paper proposes a new processor architecture called VVSHP for accelerating data-parallel applications, which are growing in importance and demanding increased performance from hardware. VVSHP merges VLIW and vector processing techniques for a simple, high-performance processor architecture. One key point of VVSHP is the execution of multiple scalar instructions within VLIW and vector instructions on unified parallel execution datapaths. Another key point is to reduce the complexity of VVSHP by designing a two-part register file: (1) shared scalar–vector part with eight-read/four-write ports 64×32-bit registers (64 scalar or 16×4 vector registers) for storing scalar/vector data and (2) vector part with two-read/one-write ports 48 vector-registers, each stores 4×32-bit vector data. Moreover, processing vector data with lengths varying from 1 to 256 represents a key point for reducing the loop overheads. VVSHP can issue up to four scalar/vector operations in each cycle for parallel processing a set of operands and producing up to four results to be written back into VVSHP register file. However, it cannot issue more than one memory operation at a time, which loads/stores 128-bit scalar/vector data from/to data memory. The design of our proposed VVSHP processor is implemented using VHDL targeting the Xilinx FPGA Virtex-5 and its performance is evaluated.</p> </sec> </abstract> … (more)
- Is Part Of:
- Microelectronics journal. Volume 46:Issue 7(2015)
- Journal:
- Microelectronics journal
- Issue:
- Volume 46:Issue 7(2015)
- Issue Display:
- Volume 46, Issue 7 (2015)
- Year:
- 2015
- Volume:
- 46
- Issue:
- 7
- Issue Sort Value:
- 2015-0046-0007-0000
- Page Start:
- 637
- Page End:
- 655
- Publication Date:
- 2015-07
- Subjects:
- Microelectronics -- Periodicals
Microélectronique -- Périodiques
Microelectronics
Electronic journals
Journals - contents and abstracts
Periodicals
621.3805 - Journal URLs:
- http://catalog.hathitrust.org/api/volumes/oclc/5877621.html ↗
http://www.sciencedirect.com/science/journal/00262692 ↗
http://www.intute.ac.uk/sciences/cgi-bin/fullrecord.pl?handle=lesa.1012319367 ↗
http://www.elsevier.com/journals ↗
http://www.elsevier.com/homepage/elecserv.htt ↗ - DOI:
- 10.1016/j.mejo.2015.03.012 ↗
- Languages:
- English
- ISSNs:
- 0959-8324
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 5758.973000
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 3783.xml