I just saw that LSI Corporation bought Sandforce, maker of high-performance SSD controllers, earlier this week for $400 million.
LSI makes a lot of things. I’ve owned a couple of SCSI controllers over the years with their chips on them. I’ve administered servers with their RAID controllers in them. They also make system-on-a-chip solutions.
LSI is into storage, so the acquisition makes sense on that level. And SSD controllers are a form of SoC, so it makes sense on that level too.
It’s only a matter of time before SSDs supplant spinning disks, and this lets LSI get in relatively early. Perhaps surprisingly, Sandforce was the last big independent controller player. OCZ bought Indilinx earlier this year. Intel, of course, is too big. The other players I can think of are Samsung (too big), Toshiba (too big), Marvell, and Jmicron. Jmicron might be small enough to acquire, but they’re only a player in the budget SSD space thanks to the uneven performance of their early controllers. Marvell also has a high-performance controller, but they’re a larger, older, and presumably more expensive company than Sandforce.
I like this move for LSI. Sandforce has had a good, long run as a producer of performance-oriented SSD controllers. The reliability has been a little bit uneven, but I don’t think LSI would have made the purchase if they didn’t think they could help with that. LSI is used to making chips that go into servers, where reliability isn’t optional. So I think this will be good for Sandforce, good for LSI, and good for the people who buy SSDs.