Entry Spertus:1995:ELB from sigplan1990.bib

Last update: Thu Apr 12 03:37:15 MDT 2012                Valid HTML 3.2!

Index sections

Top | Symbols | Numbers | Math | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

BibTeX entry

@Article{Spertus:1995:ELB,
  author =       "Ellen Spertus and William J. Dally",
  title =        "Evaluating the locality benefits of active messages",
  journal =      j-SIGPLAN,
  volume =       "30",
  number =       "8",
  pages =        "189--198",
  month =        aug,
  year =         "1995",
  CODEN =        "SINODQ",
  ISSN =         "0362-1340 (print), 1523-2867 (print), 1558-1160 (electronic)",
  ISSN-L =       "0362-1340",
  bibdate =      "Sun Dec 14 09:17:08 MST 2003",
  bibsource =    "http://portal.acm.org/",
  abstract =     "A major challenge in fine-grained computing is
                 achieving locality without excessive scheduling
                 overhead. We built two J-Machine implementations of a
                 fine-grained programming model, the Berkeley Threaded
                 Abstract Machine. One implementation takes an active
                 messages approach, maintaining a scheduling hierarchy
                 in software in order to improve data cache performance.
                 Another approach relies on the J-Machine's message
                 queues and fast task switch, lowering the control costs
                 at the expense of data locality. Our analysis measures
                 the costs and benefits of each approach, for a variety
                 of programs and cache configurations. The active
                 messages implementation is strongest when miss
                 penalties are high and for the finest-grained programs.
                 The hardware-buffered implementation is strongest in
                 direct-mapped caches, where it achieves substantially
                 better instruction cache performance.",
  acknowledgement = ack-nhfb,
  affiliation =  "Lab. for Comput. Sci., MIT, Cambridge, MA, USA",
  classification = "C6110P (Parallel programming); C6120 (File
                 organisation); C6150C (Compilers, interpreters and
                 other processors); C6150N (Distributed systems
                 software)",
  keywords =     "Active messages; Benefits; Berkeley Threaded Abstract
                 Machine; Cache configuration; Costs; Data cache
                 performance; Data locality; Direct-mapped caches; Fast
                 task switch; Fine-grained computing; Fine-grained
                 programming model; Hardware-buffered; Instruction cache
                 performance; J-Machine; Locality benefits; Message
                 queues; Miss penalties; Scheduling hierarchy;
                 Scheduling overhead",
  thesaurus =    "Cache storage; Cost-benefit analysis; Parallel
                 programming; Program compilers; Scheduling; Software
                 performance evaluation",
}

Related entries