Changes between Version 1 and Version 2 of SIMD/Design


Ignore:
Timestamp:
Apr 16, 2012 1:23:56 PM (3 years ago)
Author:
gmainland
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • SIMD/Design

    v1 v2  
    11[[PageOutline]]
    22
    3 = Using SIMD instructions in GHC =
    4 
    5 '''Goal''': improve program running times by taking advantage of CPU's SIMD vector instructions.
    6 
    7 '''How''': by extending GHC to generate code using SIMD vector instructions and by modifying libraries as necessary.
    8 
    9 This page describes the issues involved and a design for implementing SIMD vector support in GHC.
    10 
    11 Related pages:
    12  * Notes on the [wiki:SIMDPlan current implementation plan]
     3This is the design document for SIMD support in GHC that resulted from the October 11, 2011 meeting at GHC HQ. Please see the [wiki:SIMD top-level GHC SIMD] page for further details.
    134
    145== Introduction ==
     
    599590
    600591 * [http://perilsofparallel.blogspot.com/2008/09/larrabee-vs-nvidia-mimd-vs-simd.html Blog article about Larrabee and Nvidia, MIMD vs. SIMD]
    601  * [wiki:SimdLlvm SIMD LLVM] A previous (LLVM-specific) iteration of this SIMD proposal.
    602  * [wiki:VectorComputing VectorComputing]  A previous proposal to make use of x86 SSE in GHC.
     592 * [wiki:SIMD/Implementation/Llvm SIMD LLVM:] A previous (LLVM-specific) iteration of this SIMD proposal.
     593 * [wiki:SIMD/Implementation/Old Vector Computing:]  A previous proposal to make use of x86 SSE in GHC.