[dm-crypt] Using AES-256 Controller Card

Matthias Schniedermeyer ms at citd.de
Mon Aug 27 20:16:49 CEST 2012


On 27.08.2012 11:33, Brian J. Murrell wrote:
> On 12-08-27 11:23 AM, Milan Broz wrote:
> > 
> > It is (SATA) Chipset FDE (full disk encryption), IOW few layers below
> > dmcrypt operates.
> 
> Is there any add-on (i.e. USB perhaps) hardware that can assist dm-crypt?

Contrary to OP: AES-NI

I did the test that is descripted on this page:
http://wiki.debianforum.de/Benchmark_f%C3%BCr_Festplattenverschl%C3%BCsselung
(Extended to 10GiB instead of the described 512MiB)

My computer (Core i7 3770): does:
10735321088 bytes (11 GB) copied, 9.62266 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.71015 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.67001 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.61184 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.79301 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.96204 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.59942 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.61019 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.72707 s, 1.1 GB/s
10735321088 bytes (11 GB) copied, 9.88943 s, 1.1 GB/s

So it's between 1066.52 MiB/s and 1027.7 MiB/s

While the test is running, one core is @100%, the other 3 @50%
(The 4 HT-Cores aren't used) so there is a bottleneck somewhere and 
this isn't the theoretical maximum.





Bis denn

-- 
Real Programmers consider "what you see is what you get" to be just as 
bad a concept in Text Editors as it is in women. No, the Real Programmer
wants a "you asked for it, you got it" text editor -- complicated, 
cryptic, powerful, unforgiving, dangerous.



More information about the dm-crypt mailing list