Nvidia may have lost the next-generation Nexus 7 contract, which would've been the bulk of Tegra sales this year, and others seem to want to stay away from Tegra 4, too, for one reason (lack of LTE) or another (high power consumption), but it seems they can still win some contracts with certain manufacturers, if they offer them a good deal, or they simply want to go for performance.
Toshiba AT10-A is one of those contracts, and it's going to be one of the very first Tegra 4 tablets to appear on the market. As expected, the performance does seem to be very impressive compared to other chips out there, and it seems to only be matched by the Galaxy S4 (the international version) with its Exynos 5 Octa chip. Both of these devices scored somewhere around 28,000 points in the AnTuTu benchmark.
Besides its chip, its apparent performance in the AnTuTu benchmark, its size (10" display), and the fact that it will run Android 4.2.1 (not the latest Android 4.2.2), we don't know too much about its other specs. So far, though, Toshiba hasn't been very aggressive in the mobile market, and it has been one of those players who seemed to just want to "take a shot at it" without doing too much work.
The fact that it will be using Tegra 4 does offer it a differentiation advantage, but that only depends on how well received Tegra 4 will be by the customers, and the fact other OEM's aren't tripping over themselves to get Tegra 4 in their devices, isn't very encouraging.
I think Nvidia biggest mistake this year, was not making Tegra 4 (and Tegra 3, too, actually) a smartphone first chip. So far if a chip hasn't been popular in smartphones, then there's a very high chance it won't be popular in tablets, either. People also tend to buy a lot more smartphones, so if a chip does end up being very popular in some smartphones, that is free publicity for the tablet market, too. Nvidia is just targeting a currently much smaller market for no good reason. They have everything to win if Tegra becomes a popular chip in smartphone, and nothing to lose. Hopefully they will adapt this strategy for Tegra 5.