[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Stata (gllamm) benchmarks for different platforms?
Below is my last argument for the Apple G5. I think I have made my point and I don't want to get a reputation of a fanatic (!)
I trust that StataCorp will continue develop Stata (on all supported systems) in a user oriented direction and support innovations (like -gllamm-) that makes it such a great platform for statistical analyses and development of ideas. A special thanks to Chinh Nguyen and Alan Riley from StataCorp for taking part in the discussion. You have once again proved that StataCorp take great interest in the concerns from the user community.
> It's performance is not 27% of the Opteron... one or two programs do not
> define a processor's performance. There are situations where the G5 can be
> slower just as there are situations where it can be much faster. That's why
> benchmarks contain a whole suite of tests and include real world tests with
> applications rather than just artificial programs. Just because gllamm doe
> not perform as well on the G5 does not mean the G5 doesn't outperform other
> processors in other areas.
I have no problem with the G5 being outperformed by the competition in many areas however, I expected higher performance in -gllamm-. That was partly an emotional reaction however it was also based on logic:
1 Double precision floating point operations (FLOP) are very demanding.
2 -gllamm- makes heavy use of FLOP
3 The G5 is very fast on FLOP
If the premises should hold, the G5 should be very fast in -gllamm-. At least, if the code is properly optimized and there are no serious flaws in other parts of the system.
* For searches and help try: