Performance difference Win 7 vs Win 10? - 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: Performance difference Win 7 vs Win 10? (/thread-6874.html) |
Performance difference Win 7 vs Win 10? - rsberzerker - 09-16-2017 Somewhat by luck, I've got my hands on two systems for a short time, one Win 7 and the other Win 10. Both have a GTX 1060 FE card. But when I run hashcat 3.6.0 on both, doing MD5, I'm getting quite a bit different speeds. The Win 7 is doing 9000 - 9500 MH/s, while the Win 10 is doing about 5600 MH/s. The Win 10 system is newer. Why such a big difference? Does the OS really matter that much? Both are running the latest drivers. RE: Performance difference Win 7 vs Win 10? - philsmd - 09-16-2017 You should be more specific about what you are trying to run. ... or are you talking about just benchmarking with -b ? If you do not just run "hashcat -b", but some other types of attack, e.g. wordlist, there might be also other factors you need to consider... like HDD/SSD speed, RAM etc The operating system and operating system version in theory shouldn't matter at all... RE: Performance difference Win 7 vs Win 10? - rsberzerker - 09-16-2017 I'm running a mask attack, from different points in the same hcmask file I generated using PACK. (I deleted some from one copy then started hashcat.) All the masks are 10 characters, not that that should make a difference. RE: Performance difference Win 7 vs Win 10? - rsberzerker - 09-19-2017 I've done some more experimenting and I'm convinced there are other factors that determine speed besides the video card and hash. When I upgraded my main cracking system to the 1060 card, I took my old card, that got about 2600 MH/s (MD5) and used it to upgrade my TV computer (use it for watch shows) and continued the cracking job I was doing there. Last night I was getting about 2100 MH/s, and tonight I got, with it idle (no shows playing) about 1100 MH/s (a restore on the same run). It got me thinking about time I tried running a hashcat mask attack on my GPU and CPU at the same time. While I gained from the CPU, my GPU hash rate fell far more than what I gained from the CPU. So I killed the CPU job and my GPU job went back to it's former rate. So, while getting 1100 MH/s, I checked my CPU usage (procexp) and it was like 95% idle. The one thing I did note is the hashcat hardware monitor is showing my older card at about 37% utilization, so that partially explains the lower rate. I checked on the windows 10 1060 card computer, and it was around 50% there. I resumed the other 1060 card (windows 7) and it's about 91% utilization. So it's not resuming that's doing it. Anyone have any ideas what the problem could be? Obviously, there's something at play that can have a big effect on your hashrate besides the card, the hash type, and the attack type. |