Changeset 963 for docs


Ignore:
Timestamp:
Mar 21, 2011, 6:20:15 PM (8 years ago)
Author:
ksherdy
Message:

Minor edits.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • docs/PACT2011/00-abstract.tex

    r954 r963  
    55byte-at-a-time model, in which they process every character token in
    66the file in a sequential fashion. Unfortunately, the byte-at-time
    7 sequential model is a fundamental hindrance on performance and and in
     7sequential model is a fundamental hindrance on performance and in
    88some cases can add up 100\% overhead to the database queries
    99themselves.
     
    1818present a second generation parser, Parabix2, that fully parallelizes
    1919the parsing operations using using parallel bitlevel logic provided in
    20 modern SIMD extensions like SSE2.  We evaluate Parabix1and Parabix2
    21 against two widely-used XML parsers, Apache's Expat and IBM's Xerces
     20modern SIMD extensions like SSE2.  We evaluate Parabix1 and Parabix2
     21against two widely-used XML parsers, James Clark's Expat and Apache's Xerces
    2222on three generations of x86 machines, including the new Intel
    23 Sandybridge. We show that Parabix2's speedup is 2$\times$---8$\times$
     23Sandy Bridge. We show that Parabix2's speedup is 2$\times$---8$\times$
    2424over Expat and Xerces. Across the different Intel machine generations,
    2525Parabix rides the scalability curve of SIMD operations whose
    2626performance inherently scales better than traditional sequential
    27 thread performance. Comparing Intel's new Sandbrige core with the Core
     27thread performance. Comparing Intel's new Sandy Bridge core with the Core
    2828i3 we observed performance improvement between 20---60\% for our
    2929Parabix parsers while sequential parsers like Xerces improve by
Note: See TracChangeset for help on using the changeset viewer.