Wednesday, 6 July 2011
NVIDIA's 28nm Tegra 4 "Wayne"
So the the NVIDIA Kal-El is not even out yet and we are already getting information on the next Tegra chip, Wayne, which will come in two versions as NVIDIA plan on targetting the smartphone, tablet, netbook and notebook market. The chip will be released in December 2011 to the software developers to get ready for the Windows 8 release in October 2012.
The less powerful chip be a quad core ARM design and there are two choices: NVIDIA licensed Cortex-A15 or own in-house 64-bit architecture of which the Cortex-A15 is most likely to be used. It will have "at least 24-core" CUDA-enabled GPU, which is 10x more powerful than the current Tegra 2. It will also have a quad core clocked to at least 1.5Ghz, the final clock speeds will depend on the power consumption as the chip needs to be around in the same envelope as the Kal-El.
The other, more powerful, chip will be an 8 core CPU with 32-64 CUDA cores and this will be used to power tablets, netbooks and very low power notebooks. What ever NVIDIA are working on, it will be DX11 (or greater) compliant and have full OpenGL 4.x and OpenCL 1.x support. PhysX was also explicitly mentioned in the "we have full support for PhysX now" way.
The 8 core ARM processor is being built to take on X86 processors as well, which for Windows 8, Intel has covered with their Atom series - (22nm Silvermont) on the low end and 22nm Ivy Bridge then AMD plans 28nm 4-core Krishna and 2-core Wichita, both based on Enhanced Bobcat CPU core and a 32nm Trinity APU with dual and quad-core designs based on Enhanced Bulldozer core.