Discussion in 'TOSAH Archive' started by Berserker, Apr 21, 2005.
Been running this for over a year!
Just signed up to Folding@Home
Mainly been crunching for UD in the past but just switched over, doubtful I'll make a considerable impact to your stats but every little helps or something like that
Welcome to the new folders, and to the new boincers. Nice to see more people donating their spare cycles to a good cause--team OcUK's world dominance
(the science is important, too)
@Nickk--are you on the OcUK team? If there is a team...
Congrats on the first WU Ken!
Thank you! Took a few days though. Can't wait to see the effect of a system upgrade.
I've got 3 rigs here idling all the time, and could get another 2 up and running with very little trouble... expect plenty of activity from me in the near future
edit... I'm jealous of Ken's certificate
Used to run SETI (got a year's worth of WUs to my name) but since I ramped up the overclock my PC fails Prime in under 7 hours so I thought it wouldn't be stable enoughto do DC.
With the exception of TKC, you're probably right unfortunately. I personally consider SuperPi to be a better test of overclock stability than Prime95, so if you can't run Prime95, then I doubt you'll be able to run SuperPi.
SETI@home is one of the more overclock-friendly DC projects. If you can't run that, then you don't have much hope with climateprediction.net, folding@home or Predictor@home (all of which are more demanding than SETI).
There is but one solution to this. Back off that unstable overclock a few notches and join us.
Could you knock the overclock down a bit? If it lasts 7 hours of Prime you'll probably only have to reduce it a few Mhz on the FSB to get it totally stable. You won't notice the speed difference and you'll be able to help us.
The Prime 95 torture test runs lots of Fast Fourier Transforms (FFT's), this is also primarily what SETI does as an initial step to interrogate the data for signals. So failing Prime 95 is an indication that you may return invalid results.
Hopefully the validation procedure will spot it if you return any duff results caused by your overclock. You won't hurt the science but you won't get credit for it either.
I would try out a few of the BOINC projects, Folding@Home will almost certainly not like your system very much.
Probably the best to try is SETI since the WUs are relatively small and fast to complete, then at least if you lose some you may still get credit for others.
Ken: nice 1WU tiffy mate
Predictor has even shorter wu's, maybe the shortest. If you fail every 7 hours, you'd probably still validate at least 2/3 of the wu's you return.
Climate Prediction doesn't validate at all, as far as I know--so you might get full credit there, at the expense of potentially returning slightly off results.
If you can, drop the overclock a tad and join the team!
Slaker alert - whats the other HT doing - a dual xeon has 2 cpus and 2 ht cpu's - stop slacking and put your back into it
For all doubters - you can even get free wallpaper if you work hard enough!!
Team OcUK are in the following WORLD positions:-
SETI@Home Classic - world #2
BOINC combined (all the boinc projects added together) - 2nd
SETI@Home BOINC - world #2
Predictor@Home BOINC - world #1
LHC@Home BOINC - world #6
Einstein@Home BOINC - world #5 (but going to be stomped by the germans)
ClimatePrediction.NET BOINC - world #14
Help improve our position in your project of choice... (or we let slacky loose)
not to forget....
Folding@Home - #110
tho we should be in the top 100 in about a week
Right-click, save tbh
Signed up for predictor... The WU's are pretty short, and I get a nice little sig. What could I ask for? (a few extra crunchers maybe?)
Well, I'm still tweaking it, so I might get it stable yet... I want to be running SETI, but I don't want to return any duff results. I was even running it over the summer when my XP1700+ rig blew and I was reduced to using a Pentium 166 to write my 1st year report on!
How long do you all generally run Prime before getting bored and deciding it's stable enough for DC? I find mine is 100% stable (except for Windows Media Player 10, which I haven't been able to get working properly ever since I installed it), even when I'm doing intensive gaming and running Prime in the background. It only fails after several hours and it NEVER crashes the entire OS, it just returns a "critical wossname error".
Can anyone explain the difference between the new and the old SETIs btw?
Not been folding for a while after a reinstall.
Started again now though, so might start to make a slow impression on the table.
I would give it a try with SETI right now. Within a few days, you'll be able to see whether your wu's crash repeatedly, and if not, whether they validate. Boinc gives you the ability to follow up on all results each computer returns, so you can see whether they keep failing.
The new and old SETIs have the same goal of finding Adwhitworth. The only difference is the software: the new SETI runs within the boinc infrastructure, which also allows you to split your time with other projects. It adds a level of complication, but I like the ability to contribute to many projects.
Separate names with a comma.