Windows SMP Beta goes Open

BillytheImpaler said:
I'll say maths when you cease with the unnecessary u's that seem to crop up all over British English. Deal? :p ;)
Click to expand...
What u's?
Why can't you all speak english english instead of saying half the words wrong and then passing it off as another language ehh? :p

Aluminum indeed... and whats with color?
I hate it, having to program in languages and having to spell colour wrong as color.
Whats the difference between color and collar? They sound the same. You need the u!
 
Joe42 said:
Whats the difference between color and collar? They sound the same. You need the u!
Click to expand...
:confused: I don't pronounce them the same at all. (Color = kuh-ler or perhaps kuh-lor, Collar = kah-ler)

But that's neither here nor there. ;)

Getting back on topic, has there been any progress on the bug that causes the cores to continue to run after you Ctrl C out of it? I'm still running the version that was in private beta as I wait for my WUs to finish.
 
Last edited:
BillytheImpaler said:
:confused: I don't pronounce them the same at all. (Color = kuh-ler or perhaps kuh-lor, Collar = kah-ler)
Click to expand...
kah-ler? deary me, its koh-lar. :p
Color sounds like koh-lor, but colour sounds like koh-lur in british english, which is a pain. ;)

Anyway nvm, back ot... are the wu's the same between the linux smp and the win smp?
 
Joe42 said:
Why can't you all speak english english
Click to expand...
British English please. We don't want all these upstarts (word for the day there ;)) claiming they've not had half of europe polluting their language.

Factually correct though they would be
 
uncle_fungus said:
The only reason Linux was 64bit, was because the implementation of MPI only worked correctly in 64bit.
Click to expand...
Didn't know that. Explains a few things.

I'm currently running 32bit Linux -> VM -> 64bit Linux to get the WUs, and it's running pretty well considering. I have a second Linux box I'd like to do that to, but it's in a datacentre. I think while writing this post I might have worked how to get VMWare working on it. :D

PS - 1400-1600 PPD from the one working VMWare box. Here's hoping the second one works too.
 
Last edited:
A bit better than:

Project : 2651
Core : Gromacs SMP
Frames : 100
Credit : 1760


-- E6400 --

Min. Time / Frame : 20mn 29s - 1237.30 ppd
Avg. Time / Frame : 20mn 34s - 1232.29 ppd


-- P-D 820 --

Min. Time / Frame : 30mn 33s - 829.59 ppd
Avg. Time / Frame : 30mn 34s - 829.14 ppd



That 820 has got to go....
 
Stelly! can you by any chance mail me a copy of the client or anything? I've been reading up with the mixed results VS linux SMP but i'd still like to try it for myself but i can't even DL it, it cuts out at 60% every time :(
 
Can someone quickly remind me of the command you pass to a single-proc service client to only finish it's current WU and not start a new one?

Thanks :)
 
mikeymike said:
Stelly! can you by any chance mail me a copy of the client or anything? I've been reading up with the mixed results VS linux SMP but i'd still like to try it for myself but i can't even DL it, it cuts out at 60% every time :(
Click to expand...

Mailing now

Stelly
 
rincinDub said:
My windows smp is using standard loops, because it didnt close properly. Is this ok.
Click to expand...

You're in luck. The SMP cores (Win/Lin/OSX) override the client's optimisation detection.

This results in contradictory messages in FAHlog.txt, but in essence optimsations are ALWAYS used regardless of whether you shutdown the client correctly or not
 
You must log in or register to reply here.
Back
Top Bottom