You may be asking yourself why on earth would Intel go with another controller? My theory is that last year they stubbornly stuck to the theory that SATA 3Gb/sec was 'good enough for everyone', and with increasing pressure from the competition they had no choice but to make some last minute changes. A good example is the recent Sandy Bridge fiasco, where last minute changes resulted in some serious heartache. Now that they are shipping a SATA 6Gb/sec capable platform, the only thing missing is a product to use that throughput. I suspect the change in direction did not leave Intel's own SSD team enough lead time to finish their 6Gb/sec part, so there was no choice but to go with another chip.

Why Marvell? Most other controllers out there can't scale nearly as well as the original Intel controllers. The only exceptions to that are Marvell and SandForce. Between those two, SandForce would have been a better choice, but they are in direct and heavy competition with Intel. While Intel has only allowed rebranding under the Kingston name, SandForce is shipping chips to literally dozens of SSD makers. Above that, manufacturers shipping SandForce are doing quite well these days. Even if Intel came knocking, I suspect SandForce would have just politely declined at this point in the game.
That leaves Marvell, who until only recently only had their chips in the Crucial C300 line. They are now branching out, and they likely see the Intel move as a necessary step to put them on the map.