Hash Benchmark Much Slower Than Expected

Hash Benchmark Much Slower Than Expected - Printable Version

+- hashcat Forum (https://hashcat.net/forum)
+-- Forum: Support (https://hashcat.net/forum/forum-3.html)
+--- Forum: hashcat (https://hashcat.net/forum/forum-45.html)
+--- Thread: Hash Benchmark Much Slower Than Expected (/thread-6927.html)

Hash Benchmark Much Slower Than Expected - bpna - 10-10-2017

Hi,

I am a newbie here, so please bear with me. After previewing this post I see the substitution for my Linux distribution and it made me chuckle. Any and all advice from people who know is awesome.

I recently got a GTX1060 and set up The-Distribution-Which-Does-Not-Handle-OpenCL-Well (Kali) Linux, then followed the instructions on here to get my drivers set up. If it's important, I am also dual-booting from a previous Windows install, followed these directions to get that set up.

When I run

Code:
hashcat -b

I get the following:

Code:
hashcat (pull/1273/head) starting in benchmark mode...

* Device #2: Not a native Intel OpenCL runtime. Expect massive speed loss.

            You can use --force to override, but do not report related errors.

OpenCL Platform #1: NVIDIA Corporation

======================================

* Device #1: GeForce GTX 1060 6GB, 1518/6072 MB allocatable, 10MCU

OpenCL Platform #2: The pocl project

====================================

* Device #2: pthread-Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz, skipped.

Hashtype: MD4

Speed.Dev.#1.....:  9970.0 MH/s (67.28ms)

Hashtype: MD5

Speed.Dev.#1.....:  5768.2 MH/s (58.14ms)

Hashtype: Half MD5

Speed.Dev.#1.....:  3884.5 MH/s (86.35ms)

Hashtype: SHA1

Speed.Dev.#1.....:  2159.7 MH/s (77.65ms)

Hashtype: SHA-256

Speed.Dev.#1.....:   836.5 MH/s (50.11ms)

...

This differs significantly from what I found on this Hashcat Forum post, and I'm wondering if I've set something up wrong.

The command

Code:
clinfo | grep -i "icd loader"

shows the exact same loaders as the The-Distribution-Which-Does-Not-Handle-OpenCL-Well (The-Distribution-Which-Does-Not-Handle-OpenCL-Well (Kali)) NVIDIA Drivers Guide that I referenced above.

The command

Code:
nvidia-smi -i 0 -q

shows the attached GPU and clock speeds.

Code:
==============NVSMI LOG==============

Timestamp                           : Mon Oct  9 18:10:50 2017

Driver Version                      : 375.82

Attached GPUs                       : 1

GPU 0000:01:00.0

   Product Name                    : GeForce GTX 1060 6GB

   Product Brand                   : GeForce

   Display Mode                    : Enabled

   Display Active                  : Enabled

   Persistence Mode                : Disabled

   Accounting Mode                 : Disabled

   Accounting Mode Buffer Size     : 1920

   Driver Model

       Current                     : N/A

       Pending                     : N/A

   Serial Number                   : N/A

Code:
Max Clocks

       Graphics                    : 1974 MHz

       SM                          : 1974 MHz

       Memory                      : 4004 MHz

       Video                       : 1708 MHz

Any other output I can get to be more helpful? Thanks for any and all help. Excited to crack some hashes

RE: Hash Benchmark Much Slower Than Expected - Flomac - 10-10-2017

It seems to be almost exactly at half speed. Anything else running? Or a second thread?

RE: Hash Benchmark Much Slower Than Expected - bpna - 10-10-2017

(10-10-2017, 02:56 AM)Flomac Wrote: It seems to be almost exactly at half speed. Anything else running? Or a second thread?

This is a good insight, and I might not be checking correctly, but it appears nothing else is running. This is a fresh install of Linux on a machine set to dual-boot with Windows. Could Windows be doing anything with the card that I need to release? To confirm: I am powering on my machine, and in the GRUB bootloader choosing Linux, so I don't believe anything should be happening with Windows.

ps shows

Quote: PID TTY          TIME CMD
2220 pts/2    00:00:00 bash
2344 pts/2    00:00:00 ps

Code:
nvidia-smi -i 0 -q

shows

Quote:    Processes
        Process ID                  : 662
            Type                    : G
            Name                    : /usr/lib/xorg/Xorg
            Used GPU Memory         : 19 MiB
        Process ID                  : 996
            Type                    : G
            Name                    : /usr/lib/xorg/Xorg
            Used GPU Memory         : 79 MiB
        Process ID                  : 1077
            Type                    : G
            Name                    : /usr/bin/gnome-shell
            Used GPU Memory         : 59 MiB

RE: Hash Benchmark Much Slower Than Expected - Chick3nman - 10-10-2017

What version of Hashcat are you using?

RE: Hash Benchmark Much Slower Than Expected - undeath - 10-10-2017

@Chick3nman: -O has no influence on benchmark mode. My own 1060's results are similar to the one in the referenced benchmark (on 4.0.0-rc1).

RE: Hash Benchmark Much Slower Than Expected - Chick3nman - 10-10-2017

You are correct, as tested in the other thread, thanks! Edited

RE: Hash Benchmark Much Slower Than Expected - bpna - 10-11-2017

Hey Chick3nman, sorry for the late reply - just got home. hashcat -V shows

Quote:pull/1273/head

Would that be current?

undeath how do I achieve your level of enlightenment?

RE: Hash Benchmark Much Slower Than Expected - atom - 10-11-2017

The -O option has an influence in -b but only if -w is used

RE: Hash Benchmark Much Slower Than Expected - bpna - 10-18-2017

I just wanted to update - I wiped The-Distribution-Which-Does-Not-Handle-OpenCL-Well (Kali) and installed Ubuntu 16.04 and after getting NVIDIA drivers installed (and a brief period where I booted to a black screen and was able to get a TTY session after modifying a bootloader config file)...

I am getting the KICKASS SPEEDS my GTX 1060 is capable of!!!

hashcat (4.0.0-rc6) starting in benchmark mode...

Benchmarking uses hand-optimized kernel code by default.
You can use it in your cracking session by setting the -O option.
Note: Using optimized kernel code limits the maximum supported password length.
To disable the optimized kernel code in benchmark mode, use the -w option.

* Device #1: WARNING! Kernel exec timeout is not disabled.
This may cause "CL_OUT_OF_RESOURCES" or related errors.
To disable the timeout, see: https://hashcat.net/q/timeoutpatch
nvmlDeviceSetPowerManagementLimit(): Insufficient Permissions

OpenCL Platform #1: NVIDIA Corporation
======================================
* Device #1: GeForce GTX 1060 6GB, 1517/6071 MB allocatable, 10MCU

Benchmark relevant options:
===========================
* --optimized-kernel-enable

Hashmode: 900 - MD4

Speed.Dev.#1.....: 20345.0 MH/s (65.92ms)

Hashmode: 0 - MD5

Speed.Dev.#1.....: 11645.2 MH/s (57.57ms)

Hashmode: 5100 - Half MD5

Speed.Dev.#1.....: 7756.5 MH/s (86.47ms)

Hashmode: 100 - SHA1

Speed.Dev.#1.....: 4422.0 MH/s (75.83ms)

Hashmode: 1400 - SHA-256

Speed.Dev.#1.....: 1659.2 MH/s (50.31ms)

Hashmode: 10800 - SHA-384

Speed.Dev.#1.....: 552.7 MH/s (75.83ms)

...

Thanks everyone