BD Ripper (3D Plus) ultra slow ripping

[kacos]

Hi,

I get about 3 hours to rip a BD and 50' for a dvd while previously (before some latest updates that was below 50' for the bd and around 15-20' for a dvd). What changed?

if you look at the logs (attached) I get:
"..info: _ConvertSMVer2CoresDRV: MapSMtoCores for SM 5.0 is undefined. Default to use 192 Cores.."
for each of the 2 gpus (I have a GTX 650Ti and a GTX 750Ti on an i7 pc).
After various tries, I currently run the latest version (9.1.3.8) and the latest nvidia beta.

Thanks

[kacos]

[kacos]

I just uninstalled 9.1.3.8, cleaned folders & registry and re-installed 9.1.3.6.
Same problem again
I will try removing one of the GPUs see if it changes anything

[kacos]

Nope - removing one card didn't change anything.
Reverting to 9.1.2.5 didn't either.

FYI if it helps I found this:

"The message,

MapSMtoCores undefined SM 3.5 is undefined (please update to the latest SDK)!

is output when an app (normally one of the CUDA samples) that uses the CUDA samples framework (helper_cuda.h) is run on a device of compute capability 3.5 and the app was built with a version of the samples framework that has not been updated to cover that compute capability.
If CUDA 5 is installed, then it seems likely that the app was built with an earlier version of the SDK."

[kacos]

Am I the only one with this problem? Any help anyone?

[CBR929]

Hi kacos
Have you updated your drivers for your GPU's lately, if so have you tried reverting back to an earlier version to see if you have the same problems?
CBR929

[kacos]

Hi,

Yes I tried 3 different versions and am currently running the latest whql from nvidia.

[kacos]

Here's what boinc says, so drivers are ok

10-Apr-14 2:16:56 µµ | | CUDA: NVIDIA GPU 0: GeForce GTX 650 Ti (driver version 335.23, CUDA version 6.0, compute capability 3.0, 1024MB, 945MB available, 1425 GFLOPS peak)
10-Apr-14 2:16:56 µµ | | CUDA: NVIDIA GPU 1: GeForce GTX 750 Ti (driver version 335.23, CUDA version 6.0, compute capability 5.0, 2048MB, 1771MB available, 2233 GFLOPS peak)
10-Apr-14 2:16:56 µµ | | OpenCL: NVIDIA GPU 0: GeForce GTX 650 Ti (driver version 335.23, device version OpenCL 1.1 CUDA, 1024MB, 945MB available, 1425 GFLOPS peak)
10-Apr-14 2:16:56 µµ | | OpenCL: NVIDIA GPU 1: GeForce GTX 750 Ti (driver version 335.23, device version OpenCL 1.1 CUDA, 2048MB, 1771MB available, 2233 GFLOPS peak)
10-Apr-14 2:16:56 µµ | | Processor: 6 GenuineIntel Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz [Family 6 Model 30 Stepping 5]
10-Apr-14 2:16:56 µµ | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 popcnt syscall nx lm tm2 pbe
10-Apr-14 2:16:56 µµ | | OS: Microsoft Windows 8.1: Professional with Media Center x64 Edition, (06.03.9600.00)
10-Apr-14 2:16:56 µµ | | Memory: 7.99 GB physical, 8.80 GB virtual

PS: the problem started when I installed the GTX750Ti as a second card. Removing it, doesn't fix it though.

[kacos]

You can see the problem here:

14:26:08:127 info: main: DVDFab - Fabcheck 1,0,3,0 - Fri Mar 28 16:26:08 2014

14:26:08:141 info: GetVideoCardInfoEx: Cards = 1, Enables = 1
14:26:08:149 info: GetVideoCardInfoEx: - NVIDIA GeForce GTX 650 Ti
14:26:08:942 info: GetVideoCardInfoEx: - 6.0 - 3.0
14:26:09:794 info: gpuGetMaxGflopsGLDeviceIdDRV: device_count - 1 - cuda_api_version = 6000 -
14:26:09:803 info: gpuGetMaxGflopsGLDeviceIdDRV: bTCC - 0 - computeMode = 0 -
14:26:09:810 info: gpuGetMaxGflopsGLDeviceIdDRV: major = 3, minor = 0, clockRate = 928000
14:26:10:024 info: gpuGetMaxGflopsGLDeviceIdDRV: deviceName - GeForce GTX 650 Ti - benchmark divided by 10000 = 71270 -
14:26:10:033 info: DetectVideoCard_CUDA: bestCudaDetectInfo 1,0,71270,3,0,GeForce GTX 650 Ti
14:26:10:039 info: fab_check::check_cuda: cudaStatus = 0 , modeType = 56 , devCap = 1
14:26:10:047 info: fab_check::check_cuda: AcquireWriteBlock OK.
14:26:10:052 info: IntelCPUInfo::IsIntelMediaSDKSupport: nProcessorSignature = 106e5
14:26:10:057 info: IntelCPUInfo::IsIntelMediaSDKSupport: strCPUBrand = Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz
14:26:10:067

info: main: end

10:21:37:617 info: main: DVDFab - Fabcheck 1,0,3,0 - Mon Apr 07 13:21:37 2014

10:21:37:641 info: GetVideoCardInfoEx: Cards = 2, Enables = 2
10:21:37:649 info: GetVideoCardInfoEx: - NVIDIA GeForce GTX 750 Ti
10:21:37:657 info: GetVideoCardInfoEx: - NVIDIA GeForce GTX 650 Ti
10:21:37:975 info: GetVideoCardInfoEx: - 6.0 - 3.0
10:21:38:320 info: gpuGetMaxGflopsGLDeviceIdDRV: device_count - 2 - cuda_api_version = 6000 -
10:21:38:330 info: gpuGetMaxGflopsGLDeviceIdDRV: bTCC - 0 - computeMode = 0 -
10:21:38:338 info: gpuGetMaxGflopsGLDeviceIdDRV: bTCC - 0 - computeMode = 0 -
10:21:38:348 info: gpuGetMaxGflopsGLDeviceIdDRV: major = 3, minor = 0, clockRate = 928000
10:21:38:355 info: gpuGetMaxGflopsGLDeviceIdDRV: deviceName - GeForce GTX 650 Ti - benchmark divided by 10000 = 71270 -
10:21:38:365 info: gpuGetMaxGflopsGLDeviceIdDRV: major = 5, minor = 0, clockRate = 1163000
10:21:38:374 info: _ConvertSMVer2CoresDRV: MapSMtoCores for SM 5.0 is undefined. Default to use 192 Cores/SM

10:21:38:383 info: gpuGetMaxGflopsGLDeviceIdDRV: deviceName - GeForce GTX 750 Ti - benchmark divided by 10000 = 111648 -
10:21:38:394 info: DetectVideoCard_CUDA: bestCudaDetectInfo 1,1,111648,5,0,GeForce GTX 750 Ti
10:21:38:402 info: fab_check::check_cuda: cudaStatus = 0 , modeType = 56 , devCap = 1
10:21:38:408 info: fab_check::check_cuda: AcquireWriteBlock OK.
10:21:38:413 info: IntelCPUInfo::IsIntelMediaSDKSupport: nProcessorSignature = 106e5
10:21:38:419 info: IntelCPUInfo::IsIntelMediaSDKSupport: strCPUBrand = Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz
10:21:38:428 info: main: end

16:04:21:699

16:54:32:287

[kacos]

if it helps

10:21:37:649 info: GetVideoCardInfoEx: - NVIDIA GeForce GTX 750 Ti
------------- info: GetVideoCardInfoEx: ????
10:21:37:657 info: GetVideoCardInfoEx: - NVIDIA GeForce GTX 650 Ti
10:21:37:975 info: GetVideoCardInfoEx: - 6.0 - 3.0

It seems fab recognises the 650Ti with "CUDA version 6.0, compute capability 3.0" but not the 750Ti with "compute capability 5.0" which is also indicated by the "MapSMtoCores for SM 5.0 is undefined"

(sorry, but I can't edit my posts which appear hours later)

[90312]

This is not something your fellow users can change or fix for you. Try communicating it directly to the support staff HERE.

[kacos]

Ok, I changed the GTX750ti with a GT630 - here it works, but both cards have a lower computing capability. I will leave it like this until this is fixed (I hope soon). The whole point of GPU acceleration is to use much as possible cpu power from ALL GPU cores ..

info: main: The Fabcheck 1,0,3,2 - Thu Apr 10 17:13:33 2014

14:13:33:190 info: GetVideoCardInfoEx: Cards = 2, Enables = 2
14:13:33:199 info: GetVideoCardInfoEx: - NVIDIA GeForce GTX 650 Ti
14:13:33:206 info: GetVideoCardInfoEx: - NVIDIA GeForce GT 630
14:13:33:405 info: GetVideoCardInfoEx: - 6.0 - 3.0
14:13:33:619 info: gpuGetMaxGflopsGLDeviceIdDRV: device_count - 2 - cuda_api_version = 6000 -
14:13:33:626 info: gpuGetMaxGflopsGLDeviceIdDRV: bTCC - 0 - computeMode = 0 -
14:13:33:632 info: gpuGetMaxGflopsGLDeviceIdDRV: bTCC - 0 - computeMode = 0 -
14:13:33:638 info: gpuGetMaxGflopsGLDeviceIdDRV: major = 3, minor = 0, clockRate = 928000
14:13:33:644 info: gpuGetMaxGflopsGLDeviceIdDRV: deviceName - GeForce GTX 650 Ti - benchmark divided by 10000 = 71270 -
14:13:33:656 info: gpuGetMaxGflopsGLDeviceIdDRV: major = 2, minor = 1, clockRate = 1620000
14:13:33:664 info: gpuGetMaxGflopsGLDeviceIdDRV: deviceName - GeForce GT 630 - benchmark divided by 10000 = 15552 -
14:13:33:672 info: DetectVideoCard_CUDA: bestCudaDetectInfo 1,0,71270,3,0,GeForce GTX 650 Ti
14:13:33:679 info: fab_check::check_cuda: cudaStatus = 0 , modeType = 56 , devCap = 1
14:13:33:684 info: fab_check::check_cuda: AcquireWriteBlock OK.
14:13:33:690 info: IntelCPUInfo::IsIntelMediaSDKSupport: nProcessorSignature = 106e5
14:13:33:696 info: IntelCPUInfo::IsIntelMediaSDKSupport: strCPUBrand = Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz
14:13:33:704 info: main: end

_______________________

0m 00.53s: fab_check_gpu is running ...
0m 01.26s: fab_check_gpu run successful
0m 01.41s: info: recommand= 8248, Shrink support= 1, enable= 1, cuda_cap= 1 , cpu_count= 8
0m 01.41s: GPU check: GPUAccelerate-28735
0m 01.42s: GPU check: GPUAccelerate setting-8248, 8248
0m 01.42s: GPU check: CUDA cap-1
0m 01.42s: GPU check: Support coreavc decode-1
0m 01.42s: GPU check: Lighting shrink status-1, 1
0m 01.42s: GPU check: Graphics info 1: NVIDIA GeForce GTX 650 Ti : :
0m 01.42s: GPU check: Graphics info 2: NVIDIA GeForce GT 630 : 6.0 : 3.0

[Wilson.Wang]

Hi,

We are check the issue. CUDA 5.0 is newer, DVDFab not support yet.

Wilson

[kacos]

Hi,

We are check the issue. CUDA 5.0 is newer, DVDFab not support yet.

Wilson

Thanks for the reply Mr Wilson.

FYI, Cuda 6 is the latest version !

driver version 335.23, CUDA version 6.0, compute capability 5.0

[MoZy]

I too am having poor performance with this card (750 ti).

Is there any target date for a fix, or should I give up on this card for now?

Thanks!