[PS: I don't want to offend you, but could you call me to Akos? Thanks. :-)
/ Sorry, but it is a nice Hungarian name, as far as i know. :-) /
>Thanks for correcting me, Akos....I'm the one who should be apologising to you! I have edited the last post but I see I've also had a brain cramps elsewhere. Very sorry about the stupidity....won't happen again!... Cheers, Rog:)
PS...will try the new version when posted.
Ran U41.03 over night on my A64 3700+ and most long results came in under
2000 seconds. WOW! There have been no invalids so I am switching over to
U41.04 and will give that a try.
Akos, it's great, for the past couple of days it's been like ! I wake up, and there’s a new optimized application waiting for me under a message board. Thanks.
Dave
There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman
Tried U41.04 compared to S39L( fastest app sse on this machine of all the options) it is slower on my Prescott 3.6... Now I check speed by how much progress I get per 5 sec segment as a short test ( and has always born out on long test over several wu's) S39L =07-08 progress...U41.04 = 06-07 progress.
It would seem that Intels sse3 library takes to these opts very poorly compared to the improvements the AMD's are reporting, Is it possible akosf that something in the setup of the Intel is missing that AMD has and needs to be compensated for?
My first WU using U41.04 was in the same ballpark as the times my 2.4 Ghz P4 (Prescott) had been recording using S41.06 (and that is a big improvement relative to my brief experience w/U41.01).
Very nice improvement. :-) I switched from D41.14 to U41.04 @ AMD-X2-4400+ 1MB Cache and time cut is about 900 seconds, now crunching the long WU in approx. 2000 seconds only. ;-) Thanx, Akos!
It would seem that Intels sse3 library takes to these opts very poorly compared to the improvements the AMD's are reporting, Is it possible akosf that something in the setup of the Intel is missing that AMD has and needs to be compensated for?
I often discovered that AMD's float-point unit are more flexible than Intel's one. I hope that it will changed by the 'Core'.
Probably i will adjust S41.07 on a Northwood to get higher speed. -> S41.07i
(Sorry but i don't have SSE3 capable intel processor, but i will buy a Conroe.)
RE: PS: I don't want to
)
Hmm, in hungarian notation, what does the 'a' mean?
SCNR :^)
RE: [PS: I don't want to
)
>Thanks for correcting me, Akos....I'm the one who should be apologising to you! I have edited the last post but I see I've also had a brain cramps elsewhere. Very sorry about the stupidity....won't happen again!... Cheers, Rog:)
PS...will try the new version when posted.
Yay! :) I got the
)
Yay! :)
I got the U41.02/03 optimization to work or at least it didn't crash while crunching it, still waiting for the credit to be issued...but it looks good
Cut down my times from 5000-6000 seconds to 3700 seconds on a P4 2.4ghz
Thanks to Akos and his programming!
Human Stupidity Is Infinite...
Ran U41.03 over night on my
)
Ran U41.03 over night on my A64 3700+ and most long results came in under
2000 seconds. WOW! There have been no invalids so I am switching over to
U41.04 and will give that a try.
Akos, it's great, for the past couple of days it's been like ! I wake up, and there’s a new optimized application waiting for me under a message board. Thanks.
Dave
There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman
My first two U41.04 results,
)
My first two U41.04 results, on Pentium 4 Prescott 2.8GHz:
z1_1255 U41.04 3269s previously with S41.06 3606 - 3640s
z1_1010 U41.04 3125s previously with S41.06 3428 - 3489s
not yet validated but looking good. I make that around a *further* 10% improvement. Thanks again to Akosf. Mike
Tried U41.04 compared to
)
Tried U41.04 compared to S39L( fastest app sse on this machine of all the options) it is slower on my Prescott 3.6... Now I check speed by how much progress I get per 5 sec segment as a short test ( and has always born out on long test over several wu's) S39L =07-08 progress...U41.04 = 06-07 progress.
It would seem that Intels sse3 library takes to these opts very poorly compared to the improvements the AMD's are reporting, Is it possible akosf that something in the setup of the Intel is missing that AMD has and needs to be compensated for?
4 completed on my machine.
)
4 completed on my machine. 30-35m vs 35-39 with s41.06
My first WU using U41.04 was
)
My first WU using U41.04 was in the same ballpark as the times my 2.4 Ghz P4 (Prescott) had been recording using S41.06 (and that is a big improvement relative to my brief experience w/U41.01).
Very nice improvement. :-) I
)
Very nice improvement. :-) I switched from D41.14 to U41.04 @ AMD-X2-4400+ 1MB Cache and time cut is about 900 seconds, now crunching the long WU in approx. 2000 seconds only. ;-) Thanx, Akos!
RE: It would seem that
)
I often discovered that AMD's float-point unit are more flexible than Intel's one. I hope that it will changed by the 'Core'.
Probably i will adjust S41.07 on a Northwood to get higher speed. -> S41.07i
(Sorry but i don't have SSE3 capable intel processor, but i will buy a Conroe.)