implemented features (Re: Decision time....)

km b kmb810 at gmail.com
Thu Jun 7 16:25:04 PDT 2007


On 07 Jun 2007 13:04:21 GMT, Rahul Siddharthan <rsidd120 at gmail.com> wrote:

In your case, you're giving different architecture arguments (prescott
vs nocomo) and yet claiming it's apples-to-apples.
cause, you have to give nocona , if your machine has EM64 extension or
whatever fancy name they have given - read gcc man page.
You don't say what compilers you were using on the two platforms, or
even whether they were the same (gcc4 can sometimes be significantly
faster).
compiler is gcc-4.1. Did you mention anything at all about your
benchmarkings, environments etc.? You're a researcher, and you know
how to publish your results in paper. I am not here to publish a
paper.
(As an aside, you don't know the difference between PDF and
postscript.)
Thanks. you are intelligent.

All that apart, you still see a 25% increase in peak mflops in the
64-bit graph.  I don't see why you don't consider that significant.
That's not 50 % performance increase as you've claimied. On the other
hand you also see 25 % increase in peak peformance vice-versa as well.
Finally, you are testing on Intel and I am testing on an AMD Opteron.
It is widely believed that opterons do much better in 64-bit mode
(unfortunately, nearly all the published benchmarks are 32-bit and
basically useless).  Cray, Sun and many others in high-performance
computing have chosen opterons for that reason.
Well, publish your benchmark results then. I know nothing about how
you compare performance. However from what I can guess,
"(I said the 32-bit
binaries are 50% slower than the 64-bit binaries.  (Eg, running time
25 secs for 64-bit, 38 secs for 32-bit.)" - (is this the level of
benchmarking you do ? - i guess no).
Cheers,
kmb
--
Something is wrong up on cloud # 9!




More information about the Kernel mailing list