Shield Tablet. Source: NVIDIA.

Let's face it: NVIDIA (NVDA 4.35%) has failed to make a name for itself in the extremely competitive smartphone processor market. The gaming specialist tried its darndest to give Qualcomm (QCOM 1.62%) a run for its money, but Qualcomm has a lot more money than NVIDIA. For all intents and purposes, NVIDIA is done with smartphones.

For example, last year's Xiaomi Mi3 had a Tegra 4 variant as well as a Snapdragon 800 flavor. The updated Mi4 that was just announced is purely a Snapdragon affair. The newest Tegra K1 chips that were announced at CES earlier this year has exactly zero design wins in phones that investors know of, and losing a spot at one of the world's fastest growing smartphone makers doesn't exactly inspire confidence.

So, where else can NVIDIA put its newest K1 chips into?

Say hello to Shield Tablet
NVIDIA has announced Shield Tablet alongside a Shield wireless controller. The company is positioning the new device as the "world's most advanced tablet built for gamers." Shield Tablet naturally runs Google Android, which has been NVIDIA's primary choice as a mobile gaming platform for quite some time. NVIDIA is also selling a tablet cover that can be used to prop up the device. Buying all three of these components will cost a total of $400.

This isn't the first time that NVIDIA has launched a first-party tablet. Late last year, NVIDIA launched Tegra Note, a reference platform for OEMs. The company sells Tegra Note 7 devices directly, but also had companies like Hewlett-Packard sell rebranded variants. Shield Tablet is a combination of NVIDIA's previous tablet attempts and its Shield handheld gaming device.

Lacking additional wins at third-party OEMs, Shield Tablet represents NVIDIA's third attempt to sell more Tegra chips by launching its own devices.

What ever happened to "Tegra K1 season"?
On the last conference call, CEO Jen-Hsun Huang said, "It's about to be Tegra K1 season." Yet the K1 is lacking in the design win department. Thus far, there have only been two other devices announced sporting the K1. Xiaomi's MiPad was the first K1 device, so NVIDIA isn't completely out of Xiaomi's product portfolio. The second was Google's Project Tango. Now we can add Shield Tablet to the list.

Of these, the MiPad is likely the most promising in terms of unit volumes. Google's experimental tablet that senses environmental surroundings certainly won't move a lot of units. Given the high price tag and relatively niche market, Shield Tablet unit sales will also likely be modest. Huang knows that NVIDIA is not doing well in smartphones, which is why he wants to target other verticals at this point:

In terms of competition, there's lots of competition out there, but our focus with TK1 is automotive, gaming, and differentiated products, and customers who are looking for differentiated processors that can really bring some excitement to their devices. My expectation is TK1 is going to do great.

The Tegra segment's results have been deteriorating for quite some time, and NVIDIA has little choice but to salvage this business after investing heavily over the years.

The good news
Fortunately, Tegra is not NVIDIA's most important business. The GPU segment remains healthy and highly profitable. Tegra was always a risky proposition to begin with, as NVIDIA wanted to diversify into other markets. It was an expensive bet that didn't pay off financially for investors, but NVIDIA can still try to squeeze out a few more design wins in other market segments.

Eventually though, it might have to give up. Considering the fact that the Tegra operating losses ate through a third of the GPU segment's operating income last fiscal year, that wouldn't be such a bad outcome at this point.