Beyond3D Forum

swaaye · 11-Jun-2012, 22:05

Apparently they discovered that the Sandforce SF-2281 controller has a bug that breaks the AES 256 encryption mode. This chip has been around for a long, long time and only Intel found this bug. Interesting.

http://www.anandtech.com/show/5971/i...return-program
http://communities.intel.com/message/158716#158716

Pressure · 12-Jun-2012, 08:57

Just goes to show that they test their products to a far greater degree than anyone else.

They do mention that a new spin of the controller should rectify the problem though, so perhaps it is only Intel's implementation of the chip?

Davros · 12-Jun-2012, 09:41

Do other makers use this controller, are they offering refunds ?

Bludd · 12-Jun-2012, 11:03

Doesn't apply to me, I don't care about the built-in encryption.

Grall · 12-Jun-2012, 11:13

Quote:

Originally Posted by Pressure

perhaps it is only Intel's implementation of the chip?

Not unless Anandtech's got it entirely wrong; as I recall from reading about this last night it's a hardware bug. If it wasn't a hardware bug, a firmware update could have solved it without a need to respin the chip...

From what I understood, the encryption algorithm doesn't provide a diverse enough encryption result, and instead bunching up the data in a certain range/ranges of numbers, making the data easier to crack.

Not sure why this wasn't noticed before. Maybe it was, and sandforce shut the fuck up about it... Wouldn't be the first time really with that company. You'd think they'd analyze the results of their own hardware algorithm, beyond just making sure it doesn't cause data corruption.

swaaye · 12-Jun-2012, 17:35

Tech Report posted that other companies are planning to let people return or swap.

Apparently the AES 128 level works fine. This makes the bugged chip on par with other controllers that don't support AES 256.

pcchen · 12-Jun-2012, 18:23

Ironically, AES-256 doesn't seem to be more secure than AES-128 (because AES-192 and AES-256 use a relatively simpler key scheduling algorithm). There are attacks which work for AES-192 and AES-256 but not for AES-128.

Man from Atlantis · 13-Jun-2012, 22:11

hmm they dont recall 330 but 520, so it's intel only problem? or they did not advertise AES256 for 330 hence they dont recall it..

12-Jun-2012, 08:57	#2
Pressure Member Join Date: Mar 2004 Posts: 751	Just goes to show that they test their products to a far greater degree than anyone else. They do mention that a new spin of the controller should rectify the problem though, so perhaps it is only Intel's implementation of the chip? __________________ Never Argue With An Idiot. They'll Lower You To Their Level And Then Beat You With Experience!

12-Jun-2012, 09:41	#3
Davros Darlek ****** Join Date: Jun 2004 Posts: 9,494	Do other makers use this controller, are they offering refunds ? __________________ Guardian of the Most holy Two Terabytes of Gaming Goodness™

13-Jun-2012, 22:11	#8
Man from Atlantis Member Join Date: Jul 2010 Location: Istanbul Posts: 727	hmm they dont recall 330 but 520, so it's intel only problem? or they did not advertise AES256 for 330 hence they dont recall it.. __________________ SiS 6326 > Ti 4200 > 9800XT > 9800GT > GTX 460 Celeron 366 > Celeron 1700 > Athlon XP 2500+ > E6300 > Q9650

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

11-Jun-2012, 22:05	#1
swaaye Entirely Suboptimal Join Date: Mar 2003 Location: WI, USA Posts: 6,845	Intel offering returns on 520 SSD Apparently they discovered that the Sandforce SF-2281 controller has a bug that breaks the AES 256 encryption mode. This chip has been around for a long, long time and only Intel found this bug. Interesting. http://www.anandtech.com/show/5971/i...return-program http://communities.intel.com/message/158716#158716

12-Jun-2012, 11:03	#4
Bludd Eric the Half-a-bee Join Date: Oct 2003 Location: The cat detector van from the Ministry of Housinge Posts: 2,050	Doesn't apply to me, I don't care about the built-in encryption.

12-Jun-2012, 17:35	#6
swaaye Entirely Suboptimal Join Date: Mar 2003 Location: WI, USA Posts: 6,845	Tech Report posted that other companies are planning to let people return or swap. Apparently the AES 128 level works fine. This makes the bugged chip on par with other controllers that don't support AES 256.

12-Jun-2012, 18:23	#7
pcchen Moderator Join Date: Feb 2002 Location: Taiwan Posts: 2,347	Ironically, AES-256 doesn't seem to be more secure than AES-128 (because AES-192 and AES-256 use a relatively simpler key scheduling algorithm). There are attacks which work for AES-192 and AES-256 but not for AES-128.

Beyond3D Forum

Welcome, Unregistered.