benchmark and real work big difference
#1
Hello, i got rtx 3070 wich shows speed of 58kH\s on benchmark for 3200 bcrypt mode

While testing on real dictionary with like 5 billion strings, i am getting this output: 

Session..........: hashcat
Status...........: Running
Hash.Mode........: 3200 (bcrypt $2*$, Blowfish (Unix))
Hash.Target......: hashbcypt.txt
Time.Started.....: Sun Aug 06 03:34:54 2023 (16 mins, 41 secs)
Time.Estimated...: Sun Jan 26 09:11:08 2025 (1 year, 174 days)
Kernel.Feature...: Pure Kernel
Guess.Base.......: File (D:\mishahuesos.txt)
Guess.Queue......: 1/1 (100.00%)
Speed.#1.........:    1472 H/s (9.86ms) @ Accel:2 Loops:16 Thr:11 Vec:1
Recovered........: 3/17 (17.65%) Digests (total), 3/17 (17.65%) Digests (new), 3/17 (17.65%) Salts
Progress.........: 1783339/83286353284 (0.00%)
Rejected.........: 1207/1783339 (0.07%)
Restore.Point....: 104307/4899197252 (0.00%)
Restore.Sub.#1...: Salt:10 Amplifier:0-1 Iteration:656-672
Candidate.Engine.: Device Generator
Candidates.#1....: 24Mar2012 -> 250d338Fd1E093Cbd1AFb7FF5447dF7
Hardware.Mon.#1..: Temp: 62c Fan: 60% Util: 99% Core:1755MHz Mem:7000MHz Bus:16


Could this error message be reason of the error:

Successfully initialized the NVIDIA main driver CUDA runtime library.

Failed to initialize NVIDIA RTC library.

* Device #1: CUDA SDK Toolkit not installed or incorrectly installed.
            CUDA SDK Toolkit required for proper device support and utilization.
            Falling back to OpenCL runtime.

* 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
Reply
#2
UPD.
I have installed CUDa 12.2

Speed has increased, but a little , from 1400 to 2000h\s
But that is anyway far far away from benchmark 60k hashes
Please help
Reply
#3
And the strange thing is that, while hashcat is running, i could see in task manager, that CPU and GPU are loaded for 1-2%
GPU usually - 0%
Reply
#4
I am also entering -w 3 , -w 4 parameter. It dows not help. Speed is low, and nothing lagging
Reply
#5
On usuall md5 i ve got 9kh\s
that is so slooow
Reply
#6
Sounds like you're dealing with a number of different issues. For one, bcrypt comes in variable "costs" or "difficult" hashes, meaning your hash may be a much much slower setting than the benchmark hash, which is cost 05. This is the second number in the bcrypt hash so you can check your hashes to tell. On top of that, the benchmark process runs an ideal bruteforce, which is a mask attack with good workload. Other attacks may have less than ideal workloads and this show less than ideal speeds.
Reply
#7
(08-06-2023, 04:14 AM)Chick3nman Wrote: Sounds like you're dealing with a number of different issues. For one, bcrypt comes in variable "costs" or "difficult" hashes, meaning your hash may be a much much slower setting than the benchmark hash, which is cost 05. This is the second number in the bcrypt hash so you can check your hashes to tell. On top of that, the benchmark process runs an ideal bruteforce, which is a mask attack with good workload. Other attacks may have less than ideal workloads and this show less than ideal speeds.

what real speed should i expect for worlist brcypt brutoforce. for 1 hash, in case benchmark shows 58000
I hope it is not 30 times lower like i have now

Plus, i could see in task manager that my PC almost has no load
GPU always 0-2%
CPU same

I really hope it could be faster, even though these are not ideal conditions
Reply
#8
(08-06-2023, 04:14 AM)Chick3nman Wrote: Sounds like you're dealing with a number of different issues. For one, bcrypt comes in variable "costs" or "difficult" hashes, meaning your hash may be a much much slower setting than the benchmark hash, which is cost 05. This is the second number in the bcrypt hash so you can check your hashes to tell. On top of that, the benchmark process runs an ideal bruteforce, which is a mask attack with good workload. Other attacks may have less than ideal workloads and this show less than ideal speeds.

second number in hash is "2"
Reply
#9
hashcat.exe -m 0 -a 3 example0.hash ?b?b?b?b?b?b?b -w 4
for this command which should have give 60% of GPU benchmark speed, i am getting 15 000 mh\s, however it should give 60+% of benchmark speed, expecting about 22-24000 mh\s .

Could you help me to spped up bcrypt hashmode somehow?
Reply
#10
Could someone assist me please? Could I boost bcrypt wordlist method somehow?
Reply