I saw and heard a couple of interesting bits about Intel and SSDs today.
First is that Intel’s new 510 series SSDs are using a Marvell controller, rather than something of Intel’s own design. I’ve seen speculation on both extremes: Intel didn’t want to be in the controller business anyway and finally there’s something out there that they’re satisfied with; or Intel just couldn’t get a controller capable of 6 Gb/s to market quickly enough, so they went and bought one instead.
I tend to favor the latter. I can’t imagine it’s more profitable for Intel to buy a controller rather than making one.
Marvell makes the controller in Micron/Crucial’s highly regarded RealSSD C300, which was something I didn’t realize. I thought Micron was making its own controller. Intel is using tweaked firmware to wring a little more performance out of the chip. But it still trails OCZ’s upcoming Vertex 3, which increasingly appears to be the drive to beat this year.
The other thing I heard is that Intel’s next generation motherboard chipset, due about mid-year, will allow users to place an SSD in one SATA channel and a conventional HDD in a second one, and the chipset will use the SSD to cache the HDD. That way, you can get SSD performance without sacrificing HDD capacity.
It’s exactly the same idea as a hybrid drive, but this offers you greater flexibility, allowing you to match drives based on your needs. You can make your own decisions regarding buffer capacity and how you want to balance speed, economy, and power consumption.
Some users may not feel the desire for the flexibility, but I imagine power users will appreciate it. I think Intel has a winner of an idea with this one, and expect it will become very popular.