Nvidia logo CES [aa] (2)

Nvidia is in a delicate position right now. Its bread and butter, making GPUs for the PC industry, is increasingly skimpy, as customers move en masse to mobile computing and PCs are left to gather dust on store shelves.

Between a rock and a hard place

The Santa Clara company embarked into a transition to the mobile business, yet, so far, its Tegra line of mobile systems-on-a-chip hasn’t made any headway against the giant of the industry, Qualcomm. The new Tegra 4 chip has only made it to one product so far, Nvidia’s own Shield console, while the LTE-integrated Tegra 4i is only expected in 2014.

Meanwhile, Qualcomm scores design win after design win, and with the new and powerful Snapdragon 800 SoC about to hit the market, it doesn’t look to be slowing down anytime soon.

Everything’s for sale

What can Nvidia do, faced with seemingly insurmountable odds? Play its strongest card – its graphics expertise. In a blog post, Nvidia announced that, starting with the Kepler architecture, it would license “GPU cores and visual computing patent portfolio to device manufacturers to serve the needs of a large piece of the market”.

This means that mobile SoC makers, such as Samsung, Qualcomm, or even Apple, will be able to use Nvidia GPUs in their designs, in exchange of a licensing fee.

This new business model will put Nvidia on a collision course with ARM and Imagination Technology. At the moment, most SoCs in the market feature Mali GPUs from ARM or PowerVR GPUs from Imagination. Nvidia uses its own GeForce GPUs for the Tegra SoCs, but moving forward, it’s willing to let other companies use GeForce GPUs in their designs. This means that Samsung, for instance, could pair a quad-core CPU from ARM with a GPU from Nvidia.

Nvidia says that it will start by licensing its latest generation architecture, Kepler, a low-power design suitable for a wide range of applications, from supercomputers to smartphones. According to AnandTech, the IP licensing will extend to future architectures as well. Moreover, Nvidia is also interested in licensing its LTE technology, which is based on the acquisition of the modem maker Icera. In other words, everything is for sale.

Why now?

So, why has Nvidia decided now to go down the path of licensing? One reason would be the difficulties it faces with Tegra. The Santa Clara company faces the prospect of not being able to get the Tegra line off the ground. Selling its intellectual property would bring a solid business model and a lot of potential revenue.

[quote qtext=”But more importantly, more devices will have the potential to take advantage of our investments. That means more of the planet’s users will be able to enjoy our advanced graphics technologies. ” qperson=”Nvidia blog” qsource=”” qposition=”left”]

But there’s more – by turning into an ARM or Imagination competitor, Nvidia can bring its technology to more devices, of all types. PCs may be dying, but the mobile industry is still booming, while wearable and other low-powered devices shape up as the next big thing. In a few years, we could see a Samsung flagship or even an iPhone containing a GPU designed by Nvidia, which is not something Nvidia can hope for right now with its Tegra SoCs. As for future classes of devices, the sky is the limit.

At the end of the day, the move to license its technology should give Nvidia a solid and flexible business model, that would allow it to continue working on the Tegra line without worrying about achieving immediate success.

Read comments