Posts: 6
Threads: 2
Joined: Oct 2014
Hy guys,
Just setup a VM in GCP, 8x Tesla V100 and installed Tesla's driver
I benchmark and get bit less than HALF all the other benchmark speeds.
Anyone had that or could help?
All benchmark I found online gave 70 to 75GH/s on NTLM, but I am barely getting 29 to 33 with -O and -w 3
Posts: 2,267
Threads: 16
Joined: Feb 2013
it's important to understand what exactly you are comparing with (it's not good and meaningful to compare apples to oranges).
How exatly do you get the speed ? do you use -b ? My guess is that you are not running that because you are talking about manually setting -O and -w 3 (which doesn't make much sense in --benchmark, because it's the default anyways).
The number of hashes (because of a lot of optimizations with single hashes) and the attack mode (-a 3 vs -a 0) makes also a lot of difference.
For -a 0 the number of rules is also very very important.
so I think that just talking about some numbers doesn't make sense without the most crucial information missing (attack mode, number of hashes, --benchmark vs non-benchmark etc).
Posts: 6
Threads: 2
Joined: Oct 2014
(11-22-2019, 07:36 PM)philsmd Wrote: it's important to understand what exactly you are comparing with (it's not good and meaningful to compare apples to oranges).
How exatly do you get the speed ? do you use -b ? My guess is that you are not running that because you are talking about manually setting -O and -w 3 (which doesn't make much sense in --benchmark, because it's the default anyways).
The number of hashes (because of a lot of optimizations with single hashes) and the attack mode (-a 3 vs -a 0) makes also a lot of difference.
For -a 0 the number of rules is also very very important.
so I think that just talking about some numbers doesn't make sense without the most crucial information missing (attack mode, number of hashes, --benchmark vs non-benchmark etc).
Hey man, I am talking about -b , the benchmark thingy. same as when i try it on actual hashes
Posts: 2,267
Threads: 16
Joined: Feb 2013
11-22-2019, 08:09 PM
(This post was last modified: 11-22-2019, 08:09 PM by philsmd.)
can you post the output of
?
it's also worth testing with the beta version from
https://hashcat.net/beta/ or even testing with CUDA (you need to install the CUDA SDK for that AND use the latest BETA version, at the moment of this writing)
but it's unlikely that the hashcat version (further older version can be downloaded and tested from
https://hashcat.net/hashcat) matters that much... it's much more likely that you are comparing incorrectly or the problem that you are not using a bare machine, but a virtual machine instead
Posts: 6
Threads: 2
Joined: Oct 2014
(11-22-2019, 08:09 PM)philsmd Wrote: can you post the output of
?
it's also worth testing with the beta version from https://hashcat.net/beta/ or even testing with CUDA (you need to install the CUDA SDK for that AND use the latest BETA version, at the moment of this writing)
but it's unlikely that the hashcat version (further older version can be downloaded and tested from https://hashcat.net/hashcat) matters that much... it's much more likely that you are comparing incorrectly or the problem that you are not using a bare machine, but a virtual machine instead
Code:
[/font][/size][/color]
nvmlDeviceGetFanSpeed(): Not Supported
nvmlDeviceGetFanSpeed(): Not Supported
nvmlDeviceGetFanSpeed(): Not Supported
nvmlDeviceGetFanSpeed(): Not Supported
nvmlDeviceGetFanSpeed(): Not Supported
nvmlDeviceGetFanSpeed(): Not Supported
OpenCL Platform #1: NVIDIA Corporation
======================================
* Device #1: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
* Device #2: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
* Device #3: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
* Device #4: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
* Device #5: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
* Device #6: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
* Device #7: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
* Device #8: Tesla V100-SXM2-16GB, 4040/16160 MB allocatable, 80MCU
Benchmark relevant options:
===========================
* --optimized-kernel-enable
Hashmode: 1000 - NTLM
Speed.#1.........: 56682.1 MH/s (7.62ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#2.........: 55677.0 MH/s (7.62ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#3.........: 55006.2 MH/s (7.60ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#4.........: 55459.1 MH/s (7.64ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#5.........: 56350.3 MH/s (7.63ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#6.........: 56351.1 MH/s (7.63ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#7.........: 54260.2 MH/s (7.63ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#8.........: 56107.3 MH/s (7.62ms) @ Accel:32 Loops:1024 Thr:256 Vec:2
Speed.#*.........: 445.9 GH/s
Started: Fri Nov 22 21:09:45 2019
Stopped: Fri Nov 22 21:10:12 2019
[color=#333333][size=small][font=Tahoma, Verdana, Arial, sans-serif]
