Moving from the Allwinner A31 to the nVidia Tegra 4 was a decision that wasn't taken lightly as it meant we had to scrap the motherboard that was completed and about to go into production with, and instead start over with the nVidia design. We were ready to go into production the first week of July with the old design, however, it was far inferior to what we are building with nVidia. The decision was made to make the jump when the opportunity presented itself. The A31 has issues, and to fix them you have to put a bunch of other chips around it to get to 4K. Even then, the 4K stresses the chip to the limit. The A31 supports H.264, but getting H.265 (HEVC) to run on that chip was going to be a nightmare as the CPU isn't that powerful and H.265 will bring it to its knees. In sharp contrast the Tegra 4 has no problem with 4K video on the chip and requires no external help to display it. H.265 is also quite a ways down the path to running, mostly using the GPU, but could run just on CPU if needed.
Then you look at all of the other added bonuses that you get with the Tegra processor. First, nVidia has already got the Playready digital rights management all dialed in. This is crucial as Hollywood studios are requiring this level of protection on digital movies, especially in 4K. Next, gaming. While the A31's PowerVR based GPU was nice, it doesn't hold a candle to the Tegra. Think about how well Ouya did, and that's based on the Tegra 3. Imagine every game that can run on the Ouya platform, can now run even better on the Nuvola. Next, add Miracasting to the list, so we can play video from your phone onto the TV wirelessly. Then add the ability to play your nVidia based PC games, wirelessly to your TV using the Nuvola as a gateway.
All of that combined was enough to push the decision to not go into production with the A31 based product and move to nVidia. Yes, that meant that we have a bit of a delay in getting to market, but will give us such a far superior product that it is worth it. While 4K is just now starting to appear in stores, it won't really gain traction until the early fall as retailers gear up for the holiday buying season. Even if we were in September instead of July, I think the decision would have been the same, however, we have the luxury of the summer to take a step back, move to the nVidia platform, and end up with a far superior product. This is better for the consumers, and its better for the company, and thereby the shareholders. At the end of the day the delay to market will have very little impact on the sales of the product as there is no competition out there grabbing sales at the moment.