Here's the work unit I just completed. It used S5T0004 to start, but as soon as I heard about that giving invalid results, I switched over to the stable S5S0003, so hopefully it will validate. It took about 14.5 hours to finish, and I think only about 1 hour was with the test version. BTW, my other unit done about half and half with the stock and S5T0004 applications is still pending, but given other comments here, I'm guessing it won't validate.
My next work unit is another long one. *sigh* I was going to use the latest stable S5S0007 version for this, but saw a few SSE3 optimized test versions here that my computer could use. Recent posts seem to indicate that S5T0710 gives invalid results, but S5T0709 gives valid results. It may be too early to draw conclusions on these versions though. I don't mind running test versions, but would like the results to have a good chance at validating, so I'm thinking of using the S5T0709 version next. Suggestions on what patch/version I should run for the next work unit?
If you have a CPU with SSE3, then the S5T0709 is the fastest/best version for you.
Thanks, Crunchers For More Power. That's what I thought from what I read here, but figured it wouldn't hurt to check. I'll let the S5T0709 version run all of my next unit and after it's done, I'll decide what to do next. Take it one unit at a time. :) Oh, I've just given this project a somewhat higher resource share to compensate for the increased work units. This should also help me return the next work unit sooner.
RE: Dont�think so.
)
Oh, OK. Thanx
Valid and credit
)
Valid and credit granted
http://einsteinathome.org/task/34966235
S5T0301 finished with S5S0007. Updating to newer patch now
I'm going to give letting a
)
I'm going to give letting a seti unit work and see if theres any noticeable diffrence in speed. Shouldn't be on the dual core but maybe the HT pc.
Both mixed results are
)
Both mixed results are valid:
Patch: S5T0307 + S5T0709
34468379
34468377
Speed: ca. 13 - 15 % faster.
Here's the work unit I just
)
Here's the work unit I just completed. It used S5T0004 to start, but as soon as I heard about that giving invalid results, I switched over to the stable S5S0003, so hopefully it will validate. It took about 14.5 hours to finish, and I think only about 1 hour was with the test version. BTW, my other unit done about half and half with the stock and S5T0004 applications is still pending, but given other comments here, I'm guessing it won't validate.
My next work unit is another long one. *sigh* I was going to use the latest stable S5S0007 version for this, but saw a few SSE3 optimized test versions here that my computer could use. Recent posts seem to indicate that S5T0710 gives invalid results, but S5T0709 gives valid results. It may be too early to draw conclusions on these versions though. I don't mind running test versions, but would like the results to have a good chance at validating, so I'm thinking of using the S5T0709 version next. Suggestions on what patch/version I should run for the next work unit?
If you have a CPU with SSE3,
)
If you have a CPU with SSE3, then the S5T0709 is the fastest/best version for you.
finally validated .Few
)
finally validated .Few seconds with S5T0001, rest S5T0301
http://einsteinathome.org/task/34951516
RE: please read the post 6
)
Sry thought I'd been thru them all... obvioulsy missed that one.
RE: If you have a CPU with
)
Thanks, Crunchers For More Power. That's what I thought from what I read here, but figured it wouldn't hurt to check. I'll let the S5T0709 version run all of my next unit and after it's done, I'll decide what to do next. Take it one unit at a time. :) Oh, I've just given this project a somewhat higher resource share to compensate for the increased work units. This should also help me return the next work unit sooner.
Pentium4, 2.66MHz,
)
Pentium4, 2.66MHz, WinXPHome
S5S0007
34695109 1h 21min 25sec
34699121 1h 21min 21sec
standard client
1h 41min 22sec
that means S5S007 is 20min faster for the short h1 Wu´s
S5S0007 is ~20% faster than standard client on this machine
thanks to akosf
britta