Hashcat 3.20 on NVIDIA GTX 480 and Intel Xeon X5482
#1
Hi everyone,

I am trying to run Hashcat 3.2 under Windows 10 (downloaded pre-compiled binaries) and I can't make it work with GPU and CPU at the same time. In fact, I can't make it work with CPU at all. I have NVIDIA GeForce GTX 480. It seems to work fine. I updated Nvidia driver to 376.33. But when I also try to include the CPU ( I have 2 quad core Xeon X5482) I get an error

clCreateContext(): CL_UNKNOWN_ERROR

I also don't understand why hashcat complains about the card not having shader model 5.0. I believe it is shader model 5.0. Anyway, GPU part seems to work.

I tried to install opencl runtime 16.1.1 - did not help.
I tried to install full intel SDK 6.3.0.1904 - did not help.

This is my output:

hashcat64.exe -m 2500 -w 3 -D 1,2  --gpu-temp-abort 80 -a3 test.hccap -1 test.hcchr ?1?1?1?1?1?1?1?1?1?1?1?1?1?1?1
hashcat (v3.20) starting...
* Device #2: Old CUDA chipset 2.0 detected, OpenCL performance is reduced.
             For ideal hashcat performance on NVIDIA GPU you need Shader Model 5.0 or higher
nvmlDeviceGetCurrPcieLinkWidth(): Not Supported
nvmlDeviceGetClockInfo(): Not Supported
nvmlDeviceGetClockInfo(): Not Supported
nvmlDeviceGetTemperatureThreshold(): Not Supported
nvmlDeviceGetTemperatureThreshold(): Not Supported
nvmlDeviceGetUtilizationRates(): Not Supported
OpenCL Platform #1: Intel(R) Corporation
========================================
* Device #1: Intel(R) Xeon(R) CPU           X5482  @ 3.20GHz, 2047/65533 MB allocatable, 8MCU
OpenCL Platform #2: NVIDIA Corporation
======================================
* Device #2: GeForce GTX 480, 384/1536 MB allocatable, 15MCU
Hashes: 1 digests; 1 unique digests, 1 unique salts
Bitmaps: 16 bits, 65536 entries, 0x0000ffff mask, 262144 bytes, 5/13 rotates
Applicable Optimizers:
* Zero-Byte
* Single-Hash
* Single-Salt
* Brute-Force
* Slow-Hash-SIMD
Watchdog: Temperature abort trigger set to 80c
Watchdog: Temperature retain trigger set to 75c
clCreateContext(): CL_UNKNOWN_ERROR
Started: Thu Dec 22 00:27:40 2016
Stopped: Thu Dec 22 00:27:40 2016
I:\Programs\Hashcat\hashcat-3.20\hashcat-3.20>pause

Thanks.
#2
It's an OpenCL runtime installation problem. Probably you had AMD or Intel OpenCL runtime added before. Read this and follow it 100%: https://hashcat.net/wiki/doku.php?id=fre...hould_i_do
#3
Atom,

thank you for the reply. I tried it and it did not work (same error). I did not find, however, how to select “Display”, AMD/NVidia/Intel in step 3. What should I do with Driver Fusion? There is no "Display" option to select. At least I did not find it. I will try again. Maybe I did not delete all required .dlls. I only found one OpenCL.dll in System32 and another OpenCL.dll in syswow64.

Thanks.
#4
(12-22-2016, 06:39 PM)ilya980 Wrote: Atom,

thank you for the reply. I tried it and it did not work (same error). I did not find, however, how to select “Display”, AMD/NVidia/Intel in step 3. What should I do with Driver Fusion? There is no "Display" option to select. At least I did not find it. I will try again. Maybe I did not delete all required .dlls. I only found one OpenCL.dll in System32 and another OpenCL.dll in syswow64.

Thanks.

Go to Driver Family > Display and it should be there. 

From what I understand you need to delete AMD/INTEL drivers. 

I did that and I get the same problem as the op. 

Sad the previous version worked perfect Sad
#5
Ok I did a test with 3.10 and it works perfectly (I redid the test to be sure)

Also I did all the steps Atom said in the early post.

But when I use 3.20 I get this msg :

hashcat (v3.20) starting...

* Device #1: Old CUDA chipset 3.5 detected, OpenCL performance is reduced.
For ideal hashcat performance on NVIDIA GPU you need Shader Model 5.0 or higher
nvmlDeviceGetCurrPcieLinkWidth(): Not Supported

nvmlDeviceGetClockInfo(): Not Supported

nvmlDeviceGetClockInfo(): Not Supported

nvmlDeviceGetTemperatureThreshold(): Not Supported

nvmlDeviceGetTemperatureThreshold(): Not Supported

nvmlDeviceGetUtilizationRates(): Not Supported


I have an Nvidia gtx 780 so Im pretty sure I have shader5.0 compatibility.

hope we find the issue.
#6
Ok I did some test and apparently I am getting the exact same speed for 3.10 and 3.20 so I guess the problem is that my card is getting old and maybe I should replace it with something more recent. Smile

EVOLVE OR DIE Big Grin I know what I want for xmas Big Grin
#7
I followed the steps to reinstall the driver and it did not work. So, after I installed Intel 16.1.1 it gives me the error in my 1-st post. I tried 3.10 and 3.00. They produce additional warnings on the GPU side and seem to work much slower than 3.20. But neither version works with CPU. Can it be that 16.1.1 is incompatible with my CPUs? The Intel openCL.dll files in System32 and Syswow64 are dated 7/31/2016. Are these the files that should be used?
#8
I incorrectly said 3.10 was slower. Same speed in 3.10 and 3.20. But CPU still does not work. So, I looked at all openCL.dll's that I have and the only one that enables GPU is the one created on 7/31/2016 at 14:31 with 81408 bytes. I also have other openCL.dll's and openCL32.dll and openCL64.dll installed by NVIDIA dated 12/11/2016, but if I copy them to system32 folder hashcat produces an error "Can not find an OpenCL ICD loader library" and does not run. I also have openCL.dll 10/26/2016 (same 81408 bytes) and it works, but again only with GPU.
#9
Quote:But when I use 3.20 I get this msg :

hashcat (v3.20) starting...

* Device #1: Old CUDA chipset 3.5 detected, OpenCL performance is reduced.
For ideal hashcat performance on NVIDIA GPU you need Shader Model 5.0 or higher
nvmlDeviceGetCurrPcieLinkWidth(): Not Supported

nvmlDeviceGetClockInfo(): Not Supported

nvmlDeviceGetClockInfo(): Not Supported

nvmlDeviceGetTemperatureThreshold(): Not Supported

nvmlDeviceGetTemperatureThreshold(): Not Supported

nvmlDeviceGetUtilizationRates(): Not Supported
I own a GT710 which has shader 5.0 too
https://www.techpowerup.com/gpudb/1990/geforce-gt-710 and cuda 8.
#10
Was anyone able to run hashcat successfully on Intel Xeon X54xx quad core processors? Thanks.