Changes between Version 41 and Version 42 of OpenCL_Memo


Ignore:
Timestamp:
Dec 12, 2011 10:57:59 PM (13 years ago)
Author:
nakasato
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • OpenCL_Memo

    v41 v42  
    11= OpenCL Notes = 
    22== Disable auto vectorization == 
    3 In the section 6.7.2 (187 page) in the OpenCL Specification Version: 1.1 (Revision: 33), "__attribute__((vec_type_hint(<typen>))" is described. This hint controls the autovectorizer in the compiler for OpenCL C. I tested this feature by dumping the assembly code for a kernel (using "float8" ) targeted for AVX instructions with Intel SDK (version 1.5). 
     3In the section 6.7.2 (187 page) in the OpenCL Specification Version: 1.1 (Revision: 33), "__attribute__((vec_type_hint(<typen>))" is described. This hint controls the autovectorizer in the compiler for OpenCL C. I tested this feature by dumping the assembly code for a kernel (heavily using "float8" ) targeted for AVX instructions with Intel SDK (version 1.5). 
    44 
    55|| attribute       || lines ||  comment    || 
     
    1010Without this attribute, the generated code includes two functions: (1) the code without unrolling and (2) the code over unrolled.  
    1111Due to this, the generated assembly file is large but we don't know which code is really used. 
     12 
     13In both code, the core part is  calculated with AVX instructions but the performance of the two codes is slightly different.  .... now investigating. 
     14 
    1215 
    1316== How to use "ioc" command equipped with Intel SDK. ==