NVIDIA revealed at the GPU Technology Conference that the first Project Denver based Tegra SoC is coming in 2015. Codenamed Parker, this chip will feature 64-bit ARMv8 cores along with graphic based on the next-gen Maxwell architecture. Parker will be made using 3D FinFET transistors and Huang claims the chip will offer a 100x increase in performance over the Tegra 2, but it's unclear whether he was referring to GPU performance, CPU performance, or a combination of the two.
Before Parker arrives, NVIDIA will ship Logan, a Tegra chip with 32-bit ARM cores and a GPU based on the Kepler architecture. Logan is anticipated to enter production in early 2014 and will be compatible with both CUDA 5 and OpenGL 4.3.
Huang also revealed Kayla, a new circuit board that pairs Tegra 3 with "a brand-new, low-power GPU". This will basically be the first Tegra product with CUDA support, the GPU connects to Tegra via PCI Express and will support OpenGL, CUDA 5, and PhysX. NVIDIA showed off a working card running ray-tracing in real-time, but didn't reeval how it will offer Kayla as an actual product.
Huang apparently didn't want to wait until next year to combine CUDA support with an ARM-based processor, so he tasked Nvidia's engineers with creating Kayla. This tablet-sized circuit board pairs a Tegra 3 processor with a "brand-new, low-power GPU" that may be related to the Kepler derivative planned for Logan. The graphics processor boasts programmable geometry shaders and supports OpenGL, CUDA 5, and PhysX. It attaches to the Tegra chip via PCI Express.