[mikkyo]

A significant milestone for the team was passed this weekend - 5M total credit in Predictor@home.

Good job everyone, we are way ahead of the #2 team now, leading by nearly 1.75M credits.

Not the type of folks that just sit around, the team also moved up a rank in Einstein@home.
If we move some focus there, we will start flying up the ranks at a pretty good clip, eventually breaking into the top ten.

[Shaktai]

Well, I plan to continue the Mac alpha app test on SETI with my G5, but my other 5 PC boxes will stay on Einstein for a while. That should help a bit. My iBook 600 will stay on Predictor. -- L'Alliance Francophone and BOINCSynergy seem to be trying to make a push on Predictor, but we are staying about 1000 points a day ahead of them, so we are okay there.

OneMacGuy bought two of my boxes, so I don't have quite as much crunch power, but expect to have an AMD dual core CPU by mid week so that will offset part of it, while still keeping my power consumption and the heat down. The little office at home is now tolerable and needs a lot less air conditioning.

[Shaktai]

Oh, and less we forget, Einstein has a project distributed altivec enabled client for Mac OS-X that is twice as fast as previous standard clients. Right now, G4 and G5 processors are the hot boxes for Einstein. Get the most out of your G4 or G5 and crunch Einstein. --

Some Examples of how fast:
* G5 iMac 1.6 ghz -20,500 seconds per unit
* AMD 64 3200 - 22,400 seconds per unit
* Pentium 4, 3.4 ghz - 2 units in 42,200 or21,100 per unit
* Pentium 4, 3.0 ghz - 2 units in 48,400 or 24,200 per unit.
* I estimate a dual PowerMac 2.5 ghz can do 2 units every 13,120 or 1 every 6560 seconds which is twice as fast as a Pentium D 840 dual core (3.2 ghz X 2) which is 13,345 per unit.

Second to the PowerMacs is the AMD's for best performance. Even a lowly Athlon 2400 can do a unit every 23,400 seconds.

LOOK HERE to see how PowerMacs dominate the top 20 computers on Einstein, going up against multi-processor Xeons and Opeterons.

[Shaktai]

And we are now #25 in Einstein and moving up.

[Ilgaz]

I am checking lots of distributed projects today especially with Altivec support focus.

I have neen very sad things as weather prediction project without any optimization is speaking about dropping the mac port, folding@home keeps sending tinkers to macs which takes really long time to complete.

They distribute a non optimized thing and wait for mac people to participate. I am glad Einstein project is Altivec enabled by default.

People at programming team are also interestingly blaming fortran compiler prices from IBM etc which is totally unrelated to such projects. I am sure they would get the best of the industry without paying anything as projects are NON COMMERCIAL.

After 5 hours of wondering on web, I thought I better send this rant here without breaking anyones heart

[reader50]

"weather prediction project without any optimization is speaking about dropping the mac port"

Do you have a link to this?

[Ilgaz]

Hi,

Yes. In fact that post forced me to come and rant here as I am really fed up with nobody optimising for G5 processor and they speak how great Mactel performs even daring to compare a P4 3.2 to a G5.

Here I ranted again

Well, here is URL of post:
http://climateapps2.oucs.ox.ac.uk/cp...a3504be363ccd4

and text (guy quotes but post non existent. removed by author?)
> It's very unlikely that there will be optimised versions for the Mac.
> In fact Carl has said that they may have to drop support for the Mac due to
> lack of resources.
> The two of them are hard pressed to get the next experiments up and running,
> along with trouble shooting the current Windows / Linux versions.
> The only Mac they have is a slow laptop, which apparently takes 6 months to
> complete a model for testing, and Mac users are only about 2% of the total
> systems.
>

[reader50]

The text that was quoted is present, but below the post with the quote. They must use an exotic bulletin board variant.

From the sound of it, dropping Mac support is only a possibility. However, optimization is a different matter. The laptop they mention is almost certainly a G3 laptop, so they could not test Altivec optimizations.