At this year’s Graphics Technology Conference (GTC 2013), Nvidia CEO Jen-Hsun Huang unveiled a rather ambitious and impressive roadmap for its mobile and desktop GPUs that aimed to "put the efficiency and speed of our massively-parallel GPU architectures into an ever-broader array of devices."
At a recent investor event, the company updated its mobile products strategy for the next 12 to 18 months, an approach that, unsurprisingly, is centered on the Tegra family of products.
The previously noted Kepler Mobile and its dramatically reduced power consumption will not be making an appearance in the upcoming Tegra 4 chipset (or its variants), but will instead be integrated into the “Logan,” the successor to Tegra 4 that will arrive sometime in 2014. Details on the Logan GPU are still unknown, but it is expected that it will feature a refined version of the Cortex-A15 CPU featured in the Tegra 4 and will be combined with a Tegra 3 SoC and a PCI-Express slot to create Nvidia’s new “Kayla” development product.
According to Nvidia’s roadmap, Logan will be succeeded by “Parker” in 2015, but considering that it will be the first range of chips to be built on FinFET architecture and the first iteration of the long delayed “Project Denver,” the time frame seems overly optimistic. It seems more likely that Parker will be released in 2016.
Interestingly, Nvidia has roped its CUDA architecture into its mobile product roadmap and claims that it has been an “enormous GPU growth” driver and whilst this may be technically true, we are inclined to believe that it mostly pertains to workstation products like Nvidia’s Tesla GPUs and are sceptical of its supposed impact on GeForce graphics cards.
This being said, we remain hopeful that Nvidia will find a way to increase the adoption of the CUDA standard and its accompanying PhysX effects simply because the technology is under active development and it has the potential to provide a richer gaming experience on mobile devices.