27 June 2013

The Future is Now

The various stock pumping boards (their self image is not so explicit) have been chewing on the STEC deal with Western Digital. The whole story of where SSD is going may end up being an essay here, but I'd offered up one comment, and I think it encapsulates most of the issues. Offered herewith:

Fusion-io's problem: its IP, if any, is limited to the bridge twixt PCIe and the rest of the machine. Last I looked (which has been a while, I admit), they still hadn't managed to make a bootable part.

STEC, on the other hand, has a tonne of IP in its controllers. The issue comes down to whether the future of SSD (NAND or otherwise) is in emulating HDD protocols, or simply exposing NAND to the processor. Sun, thence Oracle, led the charge some years ago with flash appliances. This, to date, has been a truly Enterprise part. But, it also appears to be the growing vector of Enterprise use of flash.

Given all of that, STEC for Western Digital looks like a play at pro-sumer, SMB, etc. where a flash array would cost more than the entirety of the IT spend. Fusion-io may have a place there, but it's still too expensive. It may end up as No Man's Land of flash storage.

No comments: