NVIDIA (NVDA 0.76%) is often asked at various investor conferences if it will be the "next TI." This is, of course, a reference to the fact that Texas Instruments (TXN -1.23%) decided to give up attempting to compete in the mobile applications processor world. The harsh truth was that the R&D required to really compete effectively in this space is staggering and without meaningful market share/revenues, competing in this business is doomed to be a money sink for investors. Indeed, NVIDIA's Tegra division has already lost over $400 million during the first three quarters of its current fiscal year.

Would bailing on Tegra even save a lot of money?
At the recent BMO Technology & Digital Media conference, an analyst boldly asked the following question,

When do [sic] you guys going to shutdown Tegra and return even more of cash to investors? How does [sic] sort of give up on this mobile effort? How does, I think you have spoken to this a little bit, but how does [sic] Jen-Hsun or you sort of react to that situation?

Interestingly, the NVIDIA representative had the following to say with respect to the major losses in Tegra,

And so, if we were to shutdown Tegra, you don't save as much money as you would think. So that's not effective as the way we allocate cost, which is perhaps in retrospect now that we're hearing these questions, first person to have asked that is -- perhaps was something we should have told that more when we told about how we allocated costs.

This is pretty interesting. While it was always pretty clear that NVIDIA did leverage a great deal of the development of its traditional GPU architectures in the development of Tegra, it's interesting to hear that the R&D leverage is actually quite a bit higher than the losses in the division would tend to suggest.

On one hand, this is great for the Tegra division as it suggests that the likelihood that NVIDIA actually exits this business (which is, frankly, the most exciting business NVIDIA has from a pure growth perspective). On the other hand, this does mean that the other businesses (which are exceptionally profitable) may not be as profitable as the numbers would suggest.

How Tegra can succeed over time
NVIDIA has been pretty upfront about the fact that its Tegra business isn't just about phones – it's about powering compute devices that benefit from a differentiated graphics processing unit. While NVIDIA is likely to gain share in the smartphone and tablet markets if it can build a track record of consistent execution with Tegra, there are plenty of other places to play.

For example, NVIDIA's automotive/in-vehicle infotainment business has about $1 billion-2 billion worth of business already booked (and should be recognized over the next several years as the cars with NVIDIA's chips ship). This is an area that NVIDIA rival, Intel, has also cited as an area of strength at its recent analyst day. On top of that, there are other areas such as point-of-sale terminals and arcade gaming machines for which Tegra would be particularly well suited.

Finally, if NVIDIA's Tegra 4i and/or its upcoming Tegra 5 turn out to be smash hits, NVIDIA could eventually gain enough momentum to breakeven and eventually turn a profit in Tegra from smartphones and tablets alone. While these markets are difficult and cutthroat (Intel and Qualcomm are the giants here), NVIDIA has done well in profitable and cutthroat businesses in the past (in the early 90's there were dozens of PC graphics chip manufacturers; today there's only a handful).

Foolish bottom line
It's becoming even clearer that NVIDIA has no intentions of becoming the "next TI" – it's going to fight tooth and nail to bring Tegra to profitability. While it's a difficult road ahead, and while the company's GPUs can (and will) be expanded into more lucrative (but more niche) areas, NVIDIA's management seems set to become a leader in the low power system-on-chip market and, in the process, a compelling growth story in the semiconductor space.