The short Einstein result was r1_0265.5_2133_S4R2a_0) and I ran it on my Gallatin (Northwood-descended 2M L3 cache--the first P4 EE), hyperthreaded, with the other BOINC job also an Einstein job:
Hi again....our 400MHz PII is about 25% through the WU and a projection of the time to completion indicates at least a 1% increase in speed over C41.00....Cheers, Rog.
RE: The short Einstein
)
Now I've run the same test result on the same host for C41.01.
C41.01 3355 seconds
So for this machine (admittedly not akosf's target architecture for this version), C41.01 slowed things down with a 1.35 cpu time ratio to C41.00.
RE: S40.12 2345
)
impossible
C41.00 -> 3486 seconds (it sounds much better)
RE: RE: S40.12 2345
)
No, not impossible, but true. The actual measurements were in minutes and seconds, so I'll report them here:
S40.12 39:05
S40.04 32:13
C41.00 41:26
C41.01 55:55
RE: No, not impossible, but
)
Could you do a new measuremnt with C41.00?
RE: Could you do a new
)
Yes, I'll start it now--results in about an hour.
I will try C41.01 on our
)
I will try C41.01 on our Pentium II.....it will take about 10 hours for a result though....Cheers, Rog
RE: RE: No, not
)
My retry was close to my previous C41.00 time
C41.00 41:11 (execution ratio to previous trial .994)
RE: RE: RE: No, not
)
Thanks. I'm really surprised.
I think a P4 with HT is no
)
I think a P4 with HT is no good basis for comparing result times of the C4x apps.
The comparison schould be done on a cpu, the C-apps are build for.
I have a Celeron 433 crunching, but it's running under WINE and many WUs are repoted with 0 time, so I will not be able to compare any results.
So let's be patient and wait for people with older CPUs reporting there experiences.
cu,
Michael
Hi again....our 400MHz PII is
)
Hi again....our 400MHz PII is about 25% through the WU and a projection of the time to completion indicates at least a 1% increase in speed over C41.00....Cheers, Rog.