hashcat Forum

I am running a GTX 1080 Founders Editiion with a +200 Mhz overclock on multiple sha1 hashes.

Can anyone share their speed for doing a similar task?  I ask because I thought this would be faster than the number that I am seeing.  I get close to 1.4 billion hashes per second.

Some numbers from hashcat follow:

Speed.Dev.#1...:  1370.6 MH/s (3105.81ms)
HWMon.Dev.#1...: Temp: 56c Fan: 75% Util:100% Core:2088Mhz Mem:4513Mhz Lanes:16

Are my clocks correct?  Core looks okay, but is Mem?

Some additional notes:
I am running headless with Ubuntu 14 using xdm (per the guide on the wiki how-to) and executed hashcat with the "-w 4" option.
When I run, "./hashcat64.bin --benchmark" with a +200 Mhz overclock, I get close to 9.3 billion hashes per second (this appears to be normal).
My next test will be to run hashcat against a single sha1 hash to see if I get close to 9.4 billion hashes per second.

*Edit*

I have confirmed that for a single hash the speed is correct.  It clocks in at around 9.3 billion hashes per second.

its very low compare to epixoip https://gist.github.com/epixoip/a83d38f4...804a270c40

You're not saying exactly how many "multiple" is, nor what attack you are running. But if you're trying to compare single hash brute force speed to e.g. multihash rule-based attacks, you simply can't do that.

(11-14-2016, 08:14 PM)epixoip Wrote: [ -> ]You're not saying exactly how many "multiple" is, nor what attack you are running. But if you're trying to compare single hash brute force speed to e.g. multihash rule-based attacks, you simply can't do that.

Multiple means millions.

I am using "-m 100 -a 3" and specifiying an hcmask file which contains strings like ?d?s?s?s?d?s?s?s?d

Running it on one hash and one hash only results in expected speeds.  Running it on millions of hashes, not so much.

I was running this with an AMD 295x2 (before it cooked the motherboard) and was getting a little over 3 billion hashes per second against sha1.  I have gone with the GTX 1080 FE due to the lower power consumption and lower heat.  I was hoping for comparable performance, but I will probably need another GPU in order to get where I was.

I wanted to rule out software settings as a result of transitioning from the 295x2 to the GTX 1080 FE.  I now realize that hashcat, or its settings, are not the culprit.  I just need more overclockable GPUs that don't cook themselves or their motherboards.

Thanks for responding.

There's a drop to expect from multihash, but with 1 million hashes I think it's still OK. Here's my results using 3.20-rc1 on a stock clocked 1080FE:

$ ./hashcat hashes -a 3 ?b?b?b?b?b?b?b -w3 --speed-only -m 100 -d 1

1 hash:

Speed.Dev.#1.....: 8826.4 MH/s (76.01ms)

1,000,000 hashes:

Speed.Dev.#1.....: 6633.7 MH/s (101.14ms)