Fråga/Varning angående SandForce

Major flaw in SandForce SSD controllers (as used in SSD 520, etc)

Hello.

I am posting this here as a warning to others to stay away from any SandForce based SSD (as used in Intel's 5xx series and others, included).

We have now had a number of catastrophic failures of these drives. It is patently apparent that SandForce couldn't design a compliant and stable SATA based SSD controller if their lives depended upon it.

The AES256 issue itself was unforgivable, but it at least it didn't lose data, although it shows well enough that such a "flaw" to pass any kind of due diligence and testing means that either there wasn't any proper testing, or worse, that they knew about it and just hoped that no one would discover it.

Now users are having major issues with the failure of the SSD's SATA controller to properly respond to ATA Sleep/Wakeup and SATA Disconnect/Reconnect events. The result is bad context restoration and corruption of the drive state information resulting in a drive locked in a "panic mode" where the controller will fail to reset and the drive will no longer appear as a SATA device to the host.

Basically, if you have an SSD that is SandForce based, you *will* lose your data at some point. It is only a matter of time.

For those shills proclaiming that "it's an incompatibility with SATA controllers", it's most definitely not - it's an incompatibility with their moron "engineers" who can't read and properly implement the SATA specification documents.

For those who will claim "mine is still working fine" all I can say is consider yourself lucky and ensure you back up your data whilst you can.

In our testing the issue is most prevalent when using SATA6G host interfaces, but it may well also occur at lower speeds.

We are in discussions with Intel and some recovery companies attempting to recover our data, but are hindered by SandForce's use of forced encryption as an attempt to obscure the data. This is not about your data security, it's really all about them trying to hinder attempts at analysis of their controllers after such failures to hide their culpability.

Not that I blame Intel directly, but I would have expected a company that values its global reputation to have performed more stringent in house testing before it committed to using and putting its name on drives that used such a flawed controller.