Re: BZDEV: New snapshot, looptest.cpp (Cray T3E,KCC)

[Todd V. <tve...@oo...>]

Looptest output for Cray T3E:

Hardware: Cray T3E (mcurie.nersc.gov), single PE, 450 MHz Alpha
Theoretical peak: 900 Mflops/PE
OS: Unicos 2.0.2.24
C++ compiler: KCC 3.2b
Backend: Cray CC 3.0.2.0
Flags: +K3 -O3 --backend -hpipeline3 --backend -hunroll --backend -haggress
       --backend -hscalar2

In-cache:
Mflops/s Description
 332.18   for, indirection, unit stride
 328.41   for, indirection, unit stride, no +=
 305.97   for, indirection, unit stride, backwards loops
  333.9    for, unroll=4, unit stride, constants loaded into temps
  333.2    for, unroll=4, unit stride, constants loaded into temps,
            no += 
 333.26    for, unroll=4, unit stride, constants loaded into temps,
        CSE for index offsets
 325.44    for, unroll=4, unit stride, constants loaded into temps,            backwards
 334.85    for, unroll=8, unit stride, constants loaded into temps
 332.27   for, indirection, unit stride, constants into temps
 332.18    for, indirection, non-unit stride
 331.64    for, indirection, non-unit stride, constants loaded into temps
  65.42    while, pointer increment, unit stride
 65.584    while, pointer increment, unit stride, 
    constants loaded into temps
  65.43    while, pointer increment, non-unit stride
 69.618    while, pointer increment, unroll=4, non-unit stride,
     constants loaded into temps
 252.13    for, unroll=4, unit stride, constants loaded into temps,            prefetching
 335.17   interlaced, for, indirection, unit stride

Out of cache:
Mflops/s Description
 55.845   for, indirection, unit stride
 55.649   for, indirection, unit stride, no +=
 22.098   for, indirection, unit stride, backwards loops
 55.726    for, unroll=4, unit stride, constants loaded into temps
 55.751    for, unroll=4, unit stride, constants loaded into temps,
            no += 
 55.725    for, unroll=4, unit stride, constants loaded into temps,
        CSE for index offsets
 21.608    for, unroll=4, unit stride, constants loaded into temps,            backwards
 55.298    for, unroll=8, unit stride, constants loaded into temps
 55.418   for, indirection, unit stride, constants into temps
 54.864    for, indirection, non-unit stride
 55.546    for, indirection, non-unit stride, constants loaded into temps
 36.809    while, pointer increment, unit stride
 37.276    while, pointer increment, unit stride, 
    constants loaded into temps
 36.869    while, pointer increment, non-unit stride
 38.936    while, pointer increment, unroll=4, non-unit stride,
     constants loaded into temps
 53.712    for, unroll=4, unit stride, constants loaded into temps,            prefetching
 33.151   interlaced, for, indirection, unit stride
--------------------- blitz-dev list --------------------------------
* To subscribe/unsubscribe: mail to maj...@oo..., with
"subscribe blitz-dev" or "unsubscribe blitz-dev" in the body of the message
* Blitz++ web page: http://oonumerics.org/blitz/