When it comes to mobile chipsets, we guess it’s safe to say that Qualcomm is probably the most dominant company. There is NVIDIA, of course, although recently we haven’t seen too many Tegra-based devices, unlike Qualcomm which seems to be everywhere. At some point, we expected to see more mid-range and entry-level handsets with the low-cost Tegra 4i chip, it hasn’t happened and probably will not happen at scale. During the quarterly earnings call, NVIDIA’s CEO said that his company will not focus on the mainstream market, but will instead aim for high-end tablets and “superphones”. Yup, not smartphones, but “superphones”. Arguably handsets like the Galaxy Note 3, Galaxy S5, LG Pro G 2 would fall into that category. Today, they all use a Qualcomm chip.
While NVIDIA’s Tegra 3 chipset did make its way into a few devices, there were not as many Tegra 4 products around. This is in part due to the competition of Asian chipset manufacturers, such as Rockchip and MediaTek who have dominated the budget scene, which NVIDIA does not plan on competing in as the company really has an edge with graphics features or performance. Even NVIDIA’s CFO, Colette Kress, acknowledged this in a document filed with US regulators stating, “The year-over-year decrease in our Tegra processor business was largely due to higher sales volume of Tegra 3 for smartphone and tablet devices in the prior year, compared with current volumes of Tegra 4.”
So aside from high-end tablets and superphones, where will the company’s focus be on? According to NVIDIA, the company will be interested in devices that will have to be able to take full advantage of the Tegra’s graphical capabilities, such as gaming systems, for instance. Huang has also revealed that NVIDIA could be interested in automobiles as well where he sees a growing demand for visual computing, such as back-seat displays and collision avoidance systems.
In the grand scheme of things, the handset market is a very difficult one, and having a fast chip or high-end graphics isn’t enough. Apple produces its own chip (SoC) and won’t buy an NVIDIA mobile chip anytime soon, and Samsung has its own line of chips as well.
Additionally, if you want to launch a phone worldwide, you need to be qualified with hundreds of carriers, and different regulatory bodies. Today, only Qualcomm offers this, and Broadcom is trying to catch up to them. Right now, NVIDIA couldn’t offer that, even if its products were performing better. Smartphones are “phones first”, and the modem is a critical piece. NVIDIA does have an LTE modem, but it will take a while before it is qualified and proven from that standpoint.
The main edge then becomes the game features and graphics performance, and SHIELD is part of a larger plan for NVIDIA to leverage its strength in gaming to create a sustainable entry point for its mobile products. Automotive vehicles are also interesting because the communications aspect can be externalized, or is secondary. Car makers also don’t mind a few dollars of extra cost and are more worried about image quality and developer support, both area where NVIDIA can compete well.