Date: Thu, 19 Dec 2019 15:36:34 +0100 From: Solar Designer <solar@...nwall.com> To: john-users@...ts.openwall.com Subject: Re: Question regarding new decryptum devices. On Thu, Dec 19, 2019 at 01:41:19PM +0100, Johny Krekan wrote: > Hello, I was quite surprised by the performances offered by the new > devices which are offered by Passware for their products: > > > https://passware.us5.list-manage.com/track/click?u=146a21ba73b9d3f2ba4b60b47&id=0e480ed878&e=023f7021ef > > What do you think of those solutions? Is there any chance that John will > support them? These look like fine GPU rigs, and should be readily supported by John. > Is there a price advantage of them against the standard GPUS? They are GPUs. The only non-standard thing is the rig assembly, with liquid cooling. I expect these to cost more, including per hash rate achieved, than a GPU rig you'd build yourself would cost. Passware got to make a profit. These GPU rigs are for government and large corporate customers, for whom they may be more cost-effective than building GPU rigs in house. Another company offering GPU rigs for password cracking is Terahash (formerly Sagitta): https://terahash.com The performance is similar to Passware's, except that the largest rig by Passware is larger (12x RTX 2080 Ti vs. 10x RTX 2080 non-Ti) and is correspondingly slightly faster (255k vs. 189k at Office 2013, per these companies' own benchmarks). In fact, it's surprising the speedup is so small, at only 35%. The higher GPU count alone is 20%, the Ti should provide a lot more on top of that, and liquid cooling was supposed to provide more - but this is somehow not seen here. Maybe Passware's software is less optimal than hashcat, or maybe the rig produces so much heat that the GPUs run at a lower clock rate even with liquid cooling. Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.