NTLM Benchmark vs Mask Attack
#1
I have 2 GTX 1080 cards, and I have run a benchmark for NTLM listed below:

Hashtype: NTLM
Speed.Dev.#1.....: 44431.8 MH/s (60.25ms)
Speed.Dev.#2.....: 44439.7 MH/s (60.07ms)
Speed.Dev.#*.....: 88871.5 MH/s


Now when I run a Mask attack my results aren't even close:

Speed.Dev.#1.....: 18905.6 MH/s (7.72ms)
Speed.Dev.#2.....: 18906.8 MH/s (7.75ms)
Speed.Dev.#*.....: 37812.4 MH/s

Command Used:
hashcat64.exe -a 3 -m 1000 hashes.txt ?a?a?a?a?a?a?a?a?a -o cracked.txt

Am i missing something to optimize my GPUs or is this expected? I am running 3.30 on Windows if it makes any difference.
#2
https://hashcat.net/wiki/doku.php?id=fre...ck_so_slow

The most likely issue: is the target hash a single hash, or multiple hashes? Multiple hashes means slower cracking.

You might also try -w 3 or -w 4 to increase the workload, but that will only increase by a few percentage points (and make your UI less usable, if that matters)
~
#3
(01-20-2017, 09:07 AM)royce Wrote: https://hashcat.net/wiki/doku.php?id=fre...ck_so_slow

The most likely issue: is the target hash a single hash, or multiple hashes? Multiple hashes means slower cracking.

You might also try -w 3 or -w 4 to increase the workload, but that will only increase by a few percentage points (and make your UI less usable, if that matters)

Hi Royce, 

Ya there is multiple hashes, 40 some odd.. I had tried -w 3 when I came across the link you provided but had minimal gains.
UI is useless anyways on this system.
#4
That looks about right, then. I get the same 50%-ish drop when I do 40 NTLM:

$ hashcat -b -m 1000
hashcat (v3.30) starting in benchmark mode...

OpenCL Platform #1: NVIDIA Corporation
======================================
* Device #1: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #2: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #3: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #4: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #5: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #6: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU

Hashtype: NTLM

Speed.Dev.#1.....: 18522.4 MH/s (94.16ms)
Speed.Dev.#2.....: 18464.8 MH/s (94.45ms)
Speed.Dev.#3.....: 18426.8 MH/s (94.65ms)
Speed.Dev.#4.....: 18565.2 MH/s (93.95ms)
Speed.Dev.#5.....: 18774.3 MH/s (92.89ms)
Speed.Dev.#6.....: 18632.2 MH/s (93.61ms)
Speed.Dev.#*.....:   111.4 GH/s

Started: Thu Jan 19 22:57:25 2017
Stopped: Thu Jan 19 22:57:34 2017

$ hashcat -m 1000 -w 4 -a 3 ntlm.hash.40 ?a?a?a?a?a?a?a?a
hashcat (v3.30) starting...

OpenCL Platform #1: NVIDIA Corporation
======================================
* Device #1: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #2: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #3: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #4: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #5: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU
* Device #6: GeForce GTX 970, 1017/4068 MB allocatable, 13MCU

Hashes: 40 digests; 40 unique digests, 1 unique salts
Bitmaps: 16 bits, 65536 entries, 0x0000ffff mask, 262144 bytes, 5/13 rotates

Applicable Optimizers:
* Zero-Byte
* Precompute-Init
* Precompute-Merkle-Demgard
* Meet-In-The-Middle
* Early-Skip
* Not-Salted
* Not-Iterated
* Single-Salt
* Brute-Force
* Raw-Hash

Watchdog: Temperature abort trigger set to 90c
Watchdog: Temperature retain trigger disabled

[s]tatus [p]ause [r]esume [b]ypass [c]heckpoint [q]uit =>


Session..........: hashtest
Status...........: Running
Hash.Type........: NTLM
Hash.Target......: ntlm.hash.40
Time.Started.....: Thu Jan 19 22:57:51 2017 (2 secs)
Time.Estimated...: Sat Jan 21 07:07:05 2017 (1 day, 8 hours)
Input.Mask.......: ?a?a?a?a?a?a?a?a [8]
Input.Queue......: 1/1 (100.00%)
Speed.Dev.#1.....:  9044.8 MH/s (385.62ms)
Speed.Dev.#2.....:  9511.5 MH/s (366.13ms)
Speed.Dev.#3.....:  9597.6 MH/s (362.82ms)
Speed.Dev.#4.....:  9684.6 MH/s (360.09ms)
Speed.Dev.#5.....:  9755.1 MH/s (357.44ms)
Speed.Dev.#6.....:  9719.2 MH/s (358.72ms)
Speed.Dev.#*.....: 57312.9 MH/s
Recovered........: 0/40 (0.00%) Digests, 0/1 (0.00%) Salts
Progress.........: 104689827840/6634204312890625 (0.00%)
Rejected.........: 0/104689827840 (0.00%)
Restore.Point....: 0/735091890625 (0.00%)
Candidates.#1....: 'Ue      -> !`jY|#
Candidates.#2....: 'UvY|#   -> !`W3z'
Candidates.#3....: 'U!3z'   -> !`emw+
Candidates.#4....: 'UzGu/   -> !`!!s3
Candidates.#5....: 'UV!s3   -> !`a[p7
Candidates.#6....: 'Uamw+   -> !`vGu/
HWMon.Dev.#1.....: Temp: 69c Fan: 20% Util:100% Core:1252Mhz Mem:3004Mhz Lanes:8
HWMon.Dev.#2.....: Temp: 64c Fan: 16% Util:100% Core:1328Mhz Mem:3004Mhz Lanes:4
HWMon.Dev.#3.....: Temp: 66c Fan: 13% Util:100% Core:1341Mhz Mem:3004Mhz Lanes:16
HWMon.Dev.#4.....: Temp: 64c Fan:  9% Util:100% Core:1354Mhz Mem:3004Mhz Lanes:4
HWMon.Dev.#5.....: Temp: 71c Fan: 17% Util:100% Core:1366Mhz Mem:3004Mhz Lanes:1
HWMon.Dev.#6.....: Temp: 71c Fan: 18% Util:100% Core:1354Mhz Mem:3004Mhz Lanes:1

[s]tatus [p]ause [r]esume [b]ypass [c]heckpoint [q]uit => ^C
~
#5
Thanks Royce.

I did one hash and got the following which is more in-line with the benchmark. Thanks for the clarification.

Speed.Dev.#1...: 41094.3 MH/s (94.51ms)
Speed.Dev.#2...: 42453.9 MH/s (93.11ms)
Speed.Dev.#*...: 83548.2 MH/s