Need POWA for ClimatePrediction!

[Rattledagger]

Uhm, seti@home now has an average of 7.233 results/wu, 12. July it was 7.01 results/wu so currently it's probably more like 10+ results per wu...


The CPDN-BOINC-version apparently has a 30% speed-increase on Intel-cpu's compared to the old version, since they're using an intel-compiler now...
On Amd there's no speed-increase.

So currently Intel has an advantage.

From my mediocre performance it seems CPDN is memory-speed & multiplier-influenced just like seti.

 

[Assimilator1]

Interesting ,nice boost for the Intel cpus!

Where did you get your info on S@H WUs from?

 

[Rattledagger]

Easy, Current total statistics and Current progress

 

[Assimilator1]

I'd already found that first one ,somehow I'd missed the 'last 24hrs' bit
The 2nd link shows no stats but does give a link to graphs showing generated WUs.However the scale is far too course to be able to get even a rough idea of how many WUs/day are generated ,not to mention ,wtf does '1.5e+08' (etc) supposed to mean anyway?:roll:.Why can't they use normal numbers?

I'm sure they used to have a page which gave a good idea of the number of times WUs were crunched ,damned if I can find it now!

I still don't see where you get your 7-10 from

 

[Rattledagger]

The 1st link shows the total number of results returned is... 1524505451
The 2nd link shows "So far, SETI@home has generated 210,214,006 workunits from the data received from Arecibo."

A quick calculation gives... 7.252 results/wu for the 5+ years seti@home has run.
The 2nd page haven't been updated the last couple of days, therefore it's a little worse now than last post.

And the 7.01 was from an earlier calculation, so the huge increase in roughly 7 weeks indicates 15+ results/wu now. :Q This since AFAIK the production has been 1.2 - 1.5 M results/day in "classic" these weeks...

 

[Assimilator1]

Got ya now

Though that doesn't mean we should let BBR pass us in the short time left

 

[Def]

Got a few machines up on BIONC CP. Still don't totally like the BIONC client, but oh well - sometimes you just gotta go with the flow.

I had almost finished one WU with CP classic, but decided I should go ahead and switch while I've got some free time.

The TeAm is doing pretty good in the overall standings too.

 

[Rattledagger]

CPDN, 03.09.2004:
Notice: Unfortunately a problem in generating workunits occurred over the past few days, so incorrect Fortran namelist files were sent. If your new run crashed immediately, this is most likely why. Please 'Reset' the project to re-'Attach' and you should get a correct workunit. Sorry for the inconvenience!