what do u mean when you refer to S40 i run an amd fx55 processor with 4 gig ram and 5.09 gig virtual ram... and hard drives (4 of them) - a very fast games machine...........
what benefits would i get from a s40 (whatever it is)????
what do u mean when you refer to S40 i run an amd fx55 processor with 4 gig ram and 5.09 gig virtual ram... and hard drives (4 of them) - a very fast games machine...........
what benefits would i get from a s40 (whatever it is)????
It's the Hotted up version of the Albert client used in Einstein. Runs good, takes no short cuts, does all the same calculations, produces all the same results and is four times faster than the old standard Albert. Bruce, the Project scientist has included many of the improvements in to the latest release. But it appears the machine code (or hex?) runs a little better on most boxes.
Bruce and Akosf are working on the fine tuning now. It's been in all the best threads ;-)
what do u mean when you refer to S40 i run an amd fx55 processor with 4 gig ram and 5.09 gig virtual ram... and hard drives (4 of them) - a very fast games machine...........
what benefits would i get from a s40 (whatever it is)????
As Mike mentioned, there have been individual threads discussing user experience with each of these as they came out.
As a round number, people are seeing Einstein@Home results complete about four times faster with the most recent akosf science applications as compared to standard one from the project.
I got a couple of of 0 credits results with S40.03 with an XP 2600+. I'm switching back to S40 and hope this will be better.
After I switched back, the next result got credits.
So far all my results with S40.03 out of the series z1_1373.5 are thrown away by the scheduler.
The new result is also out of z1_1373.5, so there seems to be a difference in accuracy between S40 and S40.03.
S40.03 is 5%-6% faster that D40 on my AMD 3500+. And the result validated! :)
You better hold your smiley. ;)
Once you might get WUs of a different frequency and then the scheduler will throw away your results.
My X2 did't have a problem so far, but they may come and the XP also crunched very well, until z1_1373.5 showed up.
You better hold your smiley. ;)
Once you might get WUs of a different frequency and then the scheduler will throw away your results.
Micha,
Thanks for the heads up! I will keep a sharper eye on the results, and if they start to not validate I will switch back to D40 (which worked very well).
what do u mean when you refer
)
what do u mean when you refer to S40 i run an amd fx55 processor with 4 gig ram and 5.09 gig virtual ram... and hard drives (4 of them) - a very fast games machine...........
what benefits would i get from a s40 (whatever it is)????
RE: what do u mean when you
)
It's the Hotted up version of the Albert client used in Einstein. Runs good, takes no short cuts, does all the same calculations, produces all the same results and is four times faster than the old standard Albert. Bruce, the Project scientist has included many of the improvements in to the latest release. But it appears the machine code (or hex?) runs a little better on most boxes.
Bruce and Akosf are working on the fine tuning now. It's been in all the best threads ;-)
Join the #1 Aussie Alliance on Einstein
RE: what do u mean when you
)
There is a sticky thread devoted to release announcements for akosf's speed-up science applications.
Optimized executables announcement thread
As Mike mentioned, there have been individual threads discussing user experience with each of these as they came out.
As a round number, people are seeing Einstein@Home results complete about four times faster with the most recent akosf science applications as compared to standard one from the project.
RE: RE: My first result
)
Completely large opera...(a german joke)
Sorry, Mr. Allen, but not very 'valid' answer.
Regards.
René
So is it the general
)
So is it the general consensus that intel/Windows boxes running S39L should be upgraded to S40.03.
I got a couple of of 0
)
I got a couple of of 0 credits results with S40.03 with an XP 2600+. I'm switching back to S40 and hope this will be better. :(
cu,
Michael
RE: I got a couple of of 0
)
After I switched back, the next result got credits.
So far all my results with S40.03 out of the series z1_1373.5 are thrown away by the scheduler.
The new result is also out of z1_1373.5, so there seems to be a difference in accuracy between S40 and S40.03.
cu,
Michael
S40.03 is 5%-6% faster that
)
S40.03 is 5%-6% faster that D40 on my AMD 3500+. And the result validated! :)
RE: S40.03 is 5%-6% faster
)
You better hold your smiley. ;)
Once you might get WUs of a different frequency and then the scheduler will throw away your results.
My X2 did't have a problem so far, but they may come and the XP also crunched very well, until z1_1373.5 showed up.
Let's see what akos is gonna answer.
cu,
micha
RE: You better hold your
)
Micha,
Thanks for the heads up! I will keep a sharper eye on the results, and if they start to not validate I will switch back to D40 (which worked very well).