Welcome to MilkyWay@home

Posts by UBT - Ben

1) Message boards : Number crunching : Disecting the new validator. (Message 38086)
Posted 5 Apr 2010 by UBT - Ben
Post:
I'm having that problem also.. :S
2) Message boards : Number crunching : Setting up s GPU's (Message 27687)
Posted 13 Jul 2009 by UBT - Ben
Post:
My 4850X2, or 2x 4850 setup, uses Cat 8.12 no probs (the 4850X2 requires Cat 8.12 from Sapphire itself, as ATIs 8.12 on its own doesn't work).

As for a dummy plug:

http://www.overclock.net/overclock-net-folding-home-team/384733-30-second-dummy-plug.html

You need the resistors. Just a DVI <-> VGA plug on its own doesn't do it.



Ah cheers. Thought that might be the problem! D'oh
3) Message boards : Number crunching : Setting up s GPU's (Message 27671)
Posted 13 Jul 2009 by UBT - Ben
Post:
That's not his problem GG - just that MW is crunching on one card only, not that BOINC is not getting enough work.

Ben:

Use Cat 8.12

Disable crossfire

Use dummy plug. (or extend the screen and make second screen primary monitor, as detailed here)

GPU-Z should show both as crossfire disabled, and both working 70%+

No special config should be needed for app_info.xml

Also, for the record, GPUs doesn't have an apostrophe, unless you are describing a feature, e.g. the GPU's colour. Same goes with CPUs, DVDs, TVs, PCs. </grammar>



I have been using Catalyst 9. something for a little while now and it seems to work fine, however if it defiantly causes problems with using dual gpus then i will roll back.

When i installed the cards i didn't have a crossfire cable to link the cards, which is OK as i have read somewhere that you couldn't use them (just like CUDA in dual configuration). GPU-Z reports crossfire is disabled.

I still wonder if the DVI -> VGA converter in the back of the 2nd card is not fooling it into thinking it has actually got a cable stuck in the back. Anyone got some good links for a dummy plug?
4) Message boards : Number crunching : Setting up s GPU's (Message 27596)
Posted 12 Jul 2009 by UBT - Ben
Post:
I have tried searching the forum for an answer to this, but as yet i haven't found anything which can help me.

Scenario:
Vista x64
2 x HD4850's (Single Cards)
Drivers installed and files copied
Latest MW application being used (v.f by Gipsel)

The problem is that i can crunch fine on the 1 card, however when the 2nd card is in, BOINC won't crunch on it (GPU-Z reports seeing both cards). I plugged in a DVI->VGA adapter into the back of the second card to hopefully fool it into thinking that a cable is plugged in (although I'm not sure if that will be enough resistance).

Is that the problem, or have i missed something which is making the 2nd GPU not work correctly?

Thanks to anyone with any input on this.


Confused cruncher...
5) Message boards : Number crunching : Milestones II (Message 25669)
Posted 16 Jun 2009 by UBT - Ben
Post:
500,000 on MW xD

Slacking! :S
6) Message boards : Number crunching : Confused..about what GPU support is coming 1st (Message 25418)
Posted 14 Jun 2009 by UBT - Ben
Post:
Just to lighten the mood a little...
If i have a cricket ball in one hand, and a cricket ball in another hand. What do i have? :-)









...A very large cricket :P
7) Message boards : Number crunching : Confused..about what GPU support is coming 1st (Message 24484)
Posted 7 Jun 2009 by UBT - Ben
Post:
Well said! :P
8) Message boards : Number crunching : Catalyst 9.4 question (Message 18839)
Posted 15 Apr 2009 by UBT - Ben
Post:
It's running fine here on the 9.4 drivers (one i had made copies of the .dll files).

Running Vista x64 SP1
9.4 Catalyst Driver
9) Message boards : Number crunching : 3rd.in - optimized apps (Message 18385)
Posted 12 Apr 2009 by UBT - Ben
Post:
A BIG thanks!! Although a bit of a pain at the moment as for some reason i can't access zslip.com to re-install the program... D'oh!
10) Message boards : Number crunching : Core i7 and Optimized App (Message 11265)
Posted 17 Feb 2009 by UBT - Ben
Post:
No problem. Nice rig! I'd like to build an i7 but with the mobo and special ram that you had to get as well it just gets very expensive, so i have opted to build another Quadcore instead :)
11) Message boards : Number crunching : Core i7 and Optimized App (Message 11258)
Posted 17 Feb 2009 by UBT - Ben
Post:
6mb DDR3 Triple Channel RAM


lol hope you have enough to run Milkyway alongside windows :-P
12) Message boards : Number crunching : Is anyone glad with the project? (Message 10154)
Posted 8 Feb 2009 by UBT - Ben
Post:
OK -- this is getting rather tedious. I think I will for now approach the problem from the client side by suspending Milkyway processing, and setting it up for 'no new work'.

Assuming we don't encounter a third run of completed workunit kill off, I'll then 'unsuspend' the project to let existing work process and clean the queues.

Then, after a couple of days of normal operations, I'll let new work flow into the workstations for processing.


Gets a bit tedious seeing the exact same post in multiple threads as well. A lot of us are frustrated, but I bet Travis, Dave and the team are more frustrated than us. After all, they ahve more than a few credits at stake!



Well said Debs!
13) Message boards : Number crunching : workunit limit increased (Message 9565)
Posted 2 Feb 2009 by UBT - Ben
Post:
Thanks Travis :)
14) Message boards : Application Code Discussion : GPU app teaser (Message 9311)
Posted 28 Jan 2009 by UBT - Ben
Post:
If someone has thought, the screenshot may be a fake, there is a video of the running app.

I've implemented a solution to the scheduling problem (at least for Windows) that seems to be working on that specific machine. It is able to calculate several WUs at once, but that doesn't raise the throughput of the app, so calculating 4 WUs at the same time on a quadcore isn't faster than one after another. It uses up one core independent on the number of concurrent WUs.
In principle it works like a small scheduler that switches between the WUs 20 times a second (HD4870, 6 times a second on a HD3850). The video shows that GPU-Z is somehow incompatible with this approach as one experiences severe lags or short freezes of the system after it is started.

Thanks Emploi for that video!



Jeeeeeezzzz... And i thought you mean't 1 result every 10 or so seconds! But 4!! Wohay, why the hell did i get Nvidia! :P

Great app!
15) Message boards : Application Code Discussion : problem with checkpoints? (Message 9236)
Posted 26 Jan 2009 by UBT - Ben
Post:
@Travis: Have you read my PM about the GPU stuff? What do you think of it?

[..]
It seems pretty cool though and your values should be within our level of tolerance for what we're doing right now. It'd be great if you could post the GPU code in here so other people can look at it as well. I'd be happy to sticky and put a stamp of approval on it after we look at it.

Nice to hear!

And I have not even told you how fast it really is. You will be surprised, I promise!

But there is still the problem that there is no support from the BOINC client for a proper scheduling of GPU tasks on ATI cards (nvidia only in the moment). But maybe one could ask the BOINC devs for that?
But as you are interested, I will go ahead and try to figure out, how it can be run also on current clients.

If there is a real chance it can be distributed as stock GPU app here, I'm definitely willing to share the GPU code. But I have to warn you, it will be a little harder to maintain than a typical C code. Unfortunately, the Stream SDK of ATI isn't that mature and some functionality is basically missing in the moment. So I had to go the hard way and actually implemented it in some kind of GPU assembler language. But as the graphics card only calculates the r loop in calculate_integral (but it is 99+% of the whole computation), one can still change the stuff around it, without the need to do anything with the GPU code.
One may ask why one should take this effort, when CUDA has matured much more already. But the simple answer is performance. ATI beats the crap out of nvidia with double precision calculations. A GTX280 is maybe on the heels of a HD3850, but a HD4870 has about three times the power.

And just as a teaser for the crowd, do you remember the days of the old 1.22 application? A single HD4870 would generate more than two times the throughput of the whole MW project at that times ;)



I'd love to see this kind of program being implemented! The only concern i have is, because a lot of people here would only be too happy to run it how would the server fair up?

With the GPU's running the WU's would either have to become much longer or the server would have to be upgraded as i don't think the network would hold up on the server end if that was the case :P
16) Message boards : Application Code Discussion : GPU app teaser (Message 9232)
Posted 26 Jan 2009 by UBT - Ben
Post:
Holy mother of moo moo's... 8seconds per WU!!

Great to see some people can develop such an application on a GPU (not the easiest of things!) Just a shame that the BOINC dev's have decided to cast ATI to one side and adopt only CUDA is all i can say.

Well done, keep it up :)
17) Message boards : Number crunching : Poll: How much credit do you think is fair? (Message 2727)
Posted 24 Mar 2008 by UBT - Ben
Post:
4




©2023 Astroinformatics Group