I use this method for self regulation plus my software attempts usually end up with failure.

You could always ask the host that prompted this request as to how he goes about this problem.

Will be some sort of app_info file well worth asking.

Michael

I am getting this error on WUs when using the optimised cuda application. i am using the latest nvidia drivers and the v6 app from Raistmer. anybody any ideas?


http://setiathome.berkeley.edu/result.php?resultid=1125118371

<core_client_version>6.6.0</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1 : GeForce 9600M GT
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1250000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 4
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 9600M GT is okay
SETI@home using CUDA accelerated device GeForce 9600M GT
Rise priority modification by Raistmer
VLAR bugfix from CUDA MB dev incorporated
Priority of worker thread rised successfully
Priority of process adjusted successfully
Total GPU memory 536543232 free GPU memory 490015232
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 0.011627
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00028 0.00000
v_ChirpData 0.01971 0.00000
v_Transpose4 0.00440 0.00000
FPU opt folding 0.01054 0.00000
Cuda error 'cudaMemcpyToSymbol(cudaAcc_GaussFit_settings, (void*) &settings, sizeof(settings))' in file 'd:/BTR/seticuda/seti_boinc/client/cuda/cudaAcc_gaussfit.cu' in line 454 : invalid device symbol.

</stderr_txt>
]]>

---

In great sadness I have to report that I uninstalled the latest Boinc Version off my PC .... Any recommendations or thoughts on this.

In your ACCOUNT-page, click on SETI@home preferences. Then click on Edit SETI@home preferences and remove the check in Use Graphics Processing Unit (GPU) if available. (can't link to the last 2 as they are personal for each account)

That way only your cpu will be used for crunching. So (relatively) safe to install BOINC again ;) Like BeemerBiker said it's more likely the cpu restarting your system, but that does point to heating problems from the GPU... as in heat can't get out of the case. Perhaps adding a fan sucking out heat will help?!

If you have more computers and want some to crunch on GPU and others not, you'll have to add separate preferences for them and assign the different computers to different preferences.

I am getting this error on WUs when using the optimised cuda application. i am using the latest nvidia drivers and the v6 app from Raistmer. anybody any ideas?


http://setiathome.berkeley.edu/result.php?resultid=1125118371

<core_client_version>6.6.0</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1 : GeForce 9600M GT
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1250000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 4
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 9600M GT is okay
SETI@home using CUDA accelerated device GeForce 9600M GT
Rise priority modification by Raistmer
VLAR bugfix from CUDA MB dev incorporated
Priority of worker thread rised successfully
Priority of process adjusted successfully
Total GPU memory 536543232 free GPU memory 490015232
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 0.011627
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00028 0.00000
v_ChirpData 0.01971 0.00000
v_Transpose4 0.00440 0.00000
FPU opt folding 0.01054 0.00000
Cuda error 'cudaMemcpyToSymbol(cudaAcc_GaussFit_settings, (void*) &settings, sizeof(settings))' in file 'd:/BTR/seticuda/seti_boinc/client/cuda/cudaAcc_gaussfit.cu' in line 454 : invalid device symbol.

</stderr_txt>
]]>

This is a VLAR... WU true angle range is : 0.011627.but I have seen that you have problems even with mid AR. I suppose you have a 181.20 version of driver!?? When you process your cash of WU, Then try to reset project. Maybe that will help!

I am getting this error on WUs when using the optimised cuda application. i am using the latest nvidia drivers and the v6 app from Raistmer. anybody any ideas?

Alex, it's the same problem you were informed of earlier in this thread and others. you have a 179.xx driver version that you haven't been able to update from, and it's not conpatable with the mod app. You need to wait for 6.08 to be released.

---

Hrmm, So I am not getting CUDA now because I optimized with the AKV8 :(

What do I need to do on an SSE2 AMD dual core processor to:
1. Run CUDA optimized
2. Run non AP optimized on the cpu (I have AP disabled in my preferences)

Thank you for explaining this for me.

What do I need to do on an SSE2 AMD dual core processor to:
1. Run CUDA optimized

Install the modified CUDA app v6 in this thread

2. Run non AP optimized on the cpu (I have AP disabled in my preferences)

Attach to another project ( such as Einstein ) in Boinc Manager.

---

SETI and only SETI for me. Sorry.

---

It's not too many computers, it's a lack of circuit breakers for this room. But we can fix it :)

A teammate use BOINC V6.6.0 and crunch simultanously MB/AP on CPU and MB on GPU. (stock apps)

He made something 'wrong' or is this possible in BOINC V6.6.0?

It would be possible also with app_info.xml ?

@ eaglescouter

I posted this earlier in this thread..

Make a test and install BOINC V6.6.0 and maybe you'll get Enhanced for CPU and GPU..

Please post your experiences.. :-)

---

For a while (last week) I was getting CUDA and CPU units (not labeled as AP) but the cpu units took many hours on a dual core Opteron 165, and about 25% of them would show 0.00% completed after many hours.

So I decided to apply an optimization. I did the one in this thread, and my CUDA units promptly started dropping as error in computation. When I deleted the optimization file from SETI, I lost all my CUDA work units and got no more.

So I uninstalled SETI And BOINC and deleted the files and folders. Reinstalled clean, applied the AKV8, and got no CUDA. I upgraded to 6.2.2 and it called for CUDA work but none is available.

So I uninstalled SETI And BOINC and deleted the files and folders. Reinstalled clean, applied the AKV8, and got no CUDA. I upgraded to 6.2.2 and it called for CUDA work but none is available.

As long as you have the AK_v8 opti cpu app installed you won't get CUDA work.

It would probably be best for you to wait till 6.08 comes out and just keeping using the AK_v8 opti app untill then. You will have a much better application at that point.

It might involve having to upgrade to Boinc 6.6.0 or 6.6.2, but then you'll have the stock setup going and you can decide where you want to go from 6.08.

---

Hi there;

In great sadness I have to report that I uninstalled the latest Boinc Version
off my PC.. It was overheating my GPU and causing my computer to shut down and restart at random. I uninstall boinc no more restarts. Also to try something else to see if my theory may be correct. I did a "Stress Test" to test my system. The program was available through the NVidia control panel. During the stress test program test the GPU as well as the CPU. During the test my PC shut down and restarted. SO sadly i took off the program after running it 3 years straight. Any recommendations or thoughts on this.

Try to get better cooler (fan) and check heat-conducting layer between GPU (CPU) and radiator, maybe it needs renewal.

I am getting this error on WUs when using the optimised cuda application. i am using the latest nvidia drivers and the v6 app from Raistmer. anybody any ideas?


http://setiathome.berkeley.edu/result.php?resultid=1125118371

<core_client_version>6.6.0</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1 : GeForce 9600M GT
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1250000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 4
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 9600M GT is okay
SETI@home using CUDA accelerated device GeForce 9600M GT
Rise priority modification by Raistmer
VLAR bugfix from CUDA MB dev incorporated
Priority of worker thread rised successfully
Priority of process adjusted successfully
Total GPU memory 536543232 free GPU memory 490015232
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 0.011627
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00028 0.00000
v_ChirpData 0.01971 0.00000
v_Transpose4 0.00440 0.00000
FPU opt folding 0.01054 0.00000
Cuda error 'cudaMemcpyToSymbol(cudaAcc_GaussFit_settings, (void*) &settings, sizeof(settings))' in file 'd:/BTR/seticuda/seti_boinc/client/cuda/cudaAcc_gaussfit.cu' in line 454 : invalid device symbol.

</stderr_txt>
]]>

What version of "latest drivers" ? If less than 180.48 try to upgrade to them (try to download some beta driver if you can;t get WHQL one for your GPU). If you can't update to 180.x - try to use stock CUDA app instead.

SETI and only SETI for me. Sorry.

If SETI only - use SETI AstroPulse opt app. AstroPulse is SETI.

Hey Raistmer...welcome back. In your absense I've taken your code to position 2 and 3 in the top hosts list. Working good.

/Anton

Hey Raistmer...welcome back. In your absense I've taken your code to position 2 and 3 in the top hosts list. Working good.

/Anton

Just great :)
What will be with CPU + GPU MB opt apps working together... Will see, future is very promising :)

Hey Raistmer...welcome back. In your absense I've taken your code to position 2 and 3 in the top hosts list. Working good.

/Anton

Your RAC looks great. I have to re-start my CUDA once again and see what it can do. BTW, just had a look at the top host. 48 CPUs, just wondering what kind of setup that would be.
______________

---

Your RAC looks great. I have to re-start my CUDA once again and see what it can do. BTW, just had a look at the top host. 48 CPUs, just wondering what kind of setup that would be.
______________

Thanks...48 cores is a bunch for sure. I feel like a midget :)

The midget-RAC is amazing though, wait a while for it to stabilize as it just might be a spike from the AP validator coming online again.

Hm.. 48 cores, Server 2008 data center edition. Sounds like a cluster of 4 quad socket quad core systems. The only problem I can think of is... I didn't think BOINC would see/use more than 16 cores?

Imagine if that rig would use all 48 cores and they ran AK_V8..

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

Raistmer,
When this new 6.6.2 comes out are you going to have to do a new package for it or can we use the V6 still?

---

PROUD MEMBER OF Team Starfire World BOINC