I'm trying something new. I'm setting the a max total, max CPU, and max GPU workunits in the config_aux.xml file. Maybe this will put a cap on things.

Right now the total max will be 16, the cpu max (per processor) will be 3, and the gpu max (per processor) will also be 3. If this seems to work, and the server can handle it we'll start to increase these limits and see how things go.

---

Checking the scheduler's log, it looks like things are limited on my end, so if anyone can verify that; that would be great. :)

---

Yes I just got a few lots of 16. I have 8 BOINC CPU cores and a HD 5970.
They all errored immediately.

looks good so far...

4/9/2011 9:07:01 PM | Milkyway@home | resumed by user
4/9/2011 9:07:02 PM | Milkyway@home | work fetch resumed by user
4/9/2011 9:07:07 PM | Milkyway@home | Sending scheduler request: To fetch work.
4/9/2011 9:07:07 PM | Milkyway@home | Requesting new tasks for ATI GPU
4/9/2011 9:07:09 PM | Milkyway@home | Scheduler request completed: got 16 new tasks
4/9/2011 9:07:11 PM | Milkyway@home | Started download of milkyway_0.57_windows_intelx86__ati14.exe
4/9/2011 9:07:11 PM | Milkyway@home | Started download of parameters-10-3s-fix-new2.txt
4/9/2011 9:07:13 PM | Milkyway@home | Finished download of parameters-10-3s-fix-new2.txt
4/9/2011 9:07:13 PM | Milkyway@home | Started download of stars-10-good.txt
4/9/2011 9:07:15 PM | Milkyway@home | Finished download of milkyway_0.57_windows_intelx86__ati14.exe
4/9/2011 9:07:15 PM | Milkyway@home | Started download of parameters-10-3s-fix2.txt
4/9/2011 9:07:16 PM | Milkyway@home | Finished download of parameters-10-3s-fix2.txt
4/9/2011 9:07:24 PM | Milkyway@home | Finished download of stars-10-good.txt
4/9/2011 9:07:24 PM | Milkyway@home | Starting de_separation_10_3s_fix20_1_29_1302397587_0
4/9/2011 9:07:24 PM | Milkyway@home | Starting task de_separation_10_3s_fix20_1_29_1302397587_0 using milkyway version 57
4/9/2011 9:08:15 PM | Milkyway@home | Sending scheduler request: To fetch work.
4/9/2011 9:08:15 PM | Milkyway@home | Requesting new tasks for ATI GPU
4/9/2011 9:08:17 PM | Milkyway@home | Scheduler request completed: got 0 new tasks
4/9/2011 9:08:17 PM | Milkyway@home | No tasks sent
4/9/2011 9:08:17 PM | Milkyway@home | This computer has reached a limit on tasks in progress

they're validating and earning 267.20 points each too (at least the first several have so far)...

...and this is with a 5870 2GB GPU in conjunction with AMD's newest official drivers (Catalyst 11.3). i'm also running BOINC v6.12.18 on a WinXP Pro 32-bit platform.

---

Yes I just got a few lots of 16. I have 8 BOINC CPU cores and a HD 5970.
They all errored immediately.

What driver version are you using? Could you try updating them?

I got 16 WUs downloaded on a 1090T & 2x5970s. BOINC Client to 0.5 days so no impediment this end for more. So your block appears to work

But .... no activity at all - been ticking away for 5 mins, no activity inside CC centre at all. Clocks running on the WU, but nothing crunching.

EDIT: Aborted them in the end. Most ran with no activity for 10-12 mins, then fell over. a few did computer error immediately. Yell if you want tests running for 5970s, I'll be around for a while

Regards
Zy

My Specs From CC Centre 11.3 on a 1090T @3.7GHz 2x5970 16Gb RAM

Driver Packaging Version 8.831.2-110308a-115928C-ATI
Catalyst Version 11.3
Provider ATI Technologies Inc.
2D Driver Version 8.01.01.1134
2D Driver File Path /REGISTRY/MACHINE/SYSTEM/ControlSet001/Control/CLASS/{4D36E968-E325-11CE-BFC1-08002BE10318}/0001
Direct3D Version 7.14.10.0817
OpenGL Version 6.14.10.10600
Catalyst Control Center Version 2011.0308.2325.42017

Primary Adapter
Graphics Card Manufacturer Powered by AMD
Graphics Chipset ATI Radeon HD 5900 Series
Device ID 689C
Vendor 1002
Subsystem ID 3020
Subsystem Vendor ID 1682
Graphics Bus Capability PCI Express 2.0
Maximum Bus Setting PCI Express 2.0 x16
BIOS Version 012.013.000.002
BIOS Part Number 113-HD5970-102
BIOS Date 2009/11/11
Memory Size 1024 MB
Memory Type GDDR5
Core Clock in MHz 725 MHz
Memory Clock in MHz 1000 MHz
Total Memory Bandwidth in GByte/s 128.0 GByte/s

Regards
Zy

I have the same problem. They all error immediately.

I suspended work for now.

If you need more info, let me know.

vista x64
catalyst driver 10.5
4870x2

It appears that version .50 on Mac OS X 10.5 works but version .31 on Mac OS X 10.4 has problems. Both machines are using Intel CPUs.

---

Plus SETI Classic = 21,082 WUs

I just noticed it is a 32-bit application in my MilkyWay directory. My Windows 7 and Catalyst drivers are 64-bit. Not sure, but possibly this would cause the errors for me.

Same with me - 32 bit MW app, but I'm setup as 64bit - dont know how this one behaves in the 32bit emulator being a GPU app.

Regards
Zy

The brook.dll file is not being sent with the ATI app, that is what is causing the error.

Maybe not, even with the file it still has a computation error.

---

They look like they're running here on an HD4770 but they use 0% GPU and just run forever with no progress...

I had done a detatch and initialisation.

There are only three files in the Project directory:

32bit app
parameters-17-3s-fix-new
stars-17

Thats it - no others

EDIT:
Extract from LogFile of the startup of the batch of 16 WUs
10/04/2011 02:11:01 | Milkyway@home | work fetch resumed by user
10/04/2011 02:11:02 | Milkyway@home | Sending scheduler request: To fetch work.
10/04/2011 02:11:02 | Milkyway@home | Requesting new tasks for ATI GPU
10/04/2011 02:11:05 | Milkyway@home | Scheduler request completed: got 16 new tasks
10/04/2011 02:11:07 | Milkyway@home | Started download of milkyway_0.57_windows_intelx86__ati14.exe
10/04/2011 02:11:07 | Milkyway@home | Started download of parameters-17-3s-fix-new.txt
10/04/2011 02:11:07 | Milkyway@home | Started download of stars-17.txt
10/04/2011 02:11:08 | Milkyway@home | Finished download of parameters-17-3s-fix-new.txt
10/04/2011 02:11:14 | Milkyway@home | Finished download of milkyway_0.57_windows_intelx86__ati14.exe
10/04/2011 02:11:25 | Milkyway@home | Finished download of stars-17.txt
10/04/2011 02:11:25 | Milkyway@home | Starting de_separation_17_3s_fix_4_180_1302397609_0


Regards
Zy

The brook.dll file is not being sent with the ATI app, that is what is causing the error.

Maybe not, even with the file it still has a computation error.

hmm...i just looked in my MW@H project folder and noticed that i don't have the brook.dll file anymore either. and yet my tasks are crunching away flawlessly and validating...

---

I have tried stock, with an app_info referencing the brook.dll file and with an app_info referencing both the brook32 and brook64 dll files and they all crash out.

So far it looks like people with 32-bit systems are running file while those of us with 64-bit systems are not.

---

I'm seeing 16 0.57 units at a time on my 4830 (no cpu asked for). I have a 4830 using the 11.3 drivers and it's cranking out a unit every 4 minutes or so. Two units have run and validated already.

---

So far it looks like people with 32-bit systems are running fine while those of us with 64-bit systems are not.

i'm very tempted to agree with you on this, except that i'm the only one with a 32-bit OS and no errors to post in this thread so far. i'm hoping some other's with 32-bit platforms can chime in and confirm that they've been getting the appropriate amount of work and that its getting validated (or at the very least not erroring out).

I'm seeing 16 0.57 units at a time on my 4830 (no cpu asked for). I have a 4830 using the 11.3 drivers and it's cranking out a unit every 4 minutes or so. Two units have run and validated already.

Allan, what OS are you running, and is it 32-bit or 64-bit?

---

Sorry forgot to add, I'm using Windows XP 32bit.

---

excellent, we're making some progress here. and while i'm tempted to say that i could very well be a 64-bit OS issue, we can't yet discount the driver version. that is, Allan and i also happen to both be using the newly released Catalyst 11.3 drivers.

---