nvidia gtx 780

Nvidia announced this week that it will begin licensing its most valuable assets: its graphics processing intellectual property. I see this as a bold (and necessary) move by Nvidia, and one that could reap major rewards.

Modern tablets and smartphones require a significant ability to process graphics for video, photos, streaming, and the like. This is a major design consideration when device manufacturers select a chip to power the device. Nvidia has been trying to create more brand awareness for its own Tegra line of high performance mobile processors and has done reasonably well, although new entries in the market (e.g., Qualcomm Snapdragon, Samsung Exynos, Intel Silvermont) are putting increasing pressure on Tegra’s perceived high end market leadership position.

Nvidia’s goal with this announcement is not to undercut its own Tegra sales. Rather, its goal is a preemptive strike against both ARM’s Mali graphics processor IP (with which it is capturing increased market share, particularly at the mid to lower end), and Imagination Technology’s graphics IP, which currently has the most graphics cores in production by a number of licensees (including Intel, and Samsung/Apple).

Nvidia clearly sees a huge potential for graphics IP licensing. A large number of smaller, but high-volume chip producers (e.g., MediaTek) are creating a large market for themselves in smartphones and tablets in the mid-to-low end of the market that are proving so popular in emerging territories. They are doing this by working with a number of OEMs in places like China and India.

But even mid-range devices now require substantial graphics processing capability. As MediaTek and others establish their positions and try to move upscale to more profitable territory, they will require a competitive product against the likes of Qualcomm’s Snapdragon. Qualcomm has its own in-house CPU and graphics technology based on a deep ARM CPU design license and its own Krait GPU IP, which it acquired some time ago from AMD. So these vendors will need to achieve at least equivalent performance, and will need to license that capability.

With this week’s announcement, Nvidia puts itself in a great position to serve the volume market for smartphones and tablets that it could not reach on its own with Tegra. By licensing its GPU’s, where it is recognized as a leader in graphics performance, it can leverage its assets to play in this fast emerging and growing market without having to make a risky investment in marketing and production partnerships. And more importantly, Nvidia won’t have to produce very low cost designs that would limit its margins dramatically. This shifts the risk to its licensees, and presents Nvidia will a very low risk posture.

Bottom Line: Nvidia’s move to license it graphics IP is essentially throwing down the gauntlet to ARM and Imagination Tech, and by extension to Qualcomm as well. I believe Nvidia can achieve a significant penetration of the emerging volume markets. With this move, Nvidia will gain both licensing revenues which could be substantial (as well as profitable). And it can then leverage the installed base to potentially up sell its own higher end Tegra chips.

The risk is that this move might ultimately impact Tegra sales, but I don’t expect this to be an issue in the short term (2-3 years). This is a winning strategy for Nvidia, and I expect to see some major partnership announcements in the near future. For the most part, it hurts Imagination Tech, which could lose licensing revenues as a result of Nvidia’s move, and to a lesser extent ARM, which generates the bulk of its revenues on the CPU IP it licenses.

Jack Gold is the founder and principal analyst at J.Gold Associates, based in Northborough, Mass. He covers the many aspects of business and consumer computing and emerging technologies.