I've been following Intel's (INTC -0.51%) mobile group for years, and for quite a while, I had been optimistic that the company would deliver world-class products in smartphones and tablets. However, here in 2015, Intel's product lineup is still pretty uninspiring. Sure, it now has an integrated applications processor and modem with the Atom x3 line, but these parts are built on relatively old, third-party manufacturing technology.

And, to be perfectly blunt, they're competent but nothing special.

The company's higher-end mobile chips -- the Atom x5 and x7 -- are built on Intel's latest 14-nanometer technology, but because these chips were late, they don't have the features or performance that anybody would call "leadership."

From the perspective of an Intel shareholder, I'm giving the company one more shot to deliver strong products that are worthy of Intel's massive spending in mobile. If by the end of 2016, Intel hasn't released products that are truly best-in-class for their respective segments, I'm going to -- for the purpose of evaluating Intel as an investment in my personal portfolio -- write off the company's mobile group.

What Intel has promised, and what I need to see
Intel has promised that in 2016, it will launch a top-to-bottom product stack for the mobile market. SoFIA LTE 2 for low-cost devices, SoFIA MID for mid-range parts, and Broxton for high-end phones.

I don't expect these parts to be massive hits overnight, but since these parts will all have the privilege of using Intel's latest 14-nanometer manufacturing technology and likely went into design around the time Intel substantially stepped up its mobile chip R&D, I expect them to be competitive. And I expect them to win designs and drive significant revenue growth at positive margins.

If Intel can deliver that, then as an investor, I can reasonably expect steady progress each and every year. This means a clear path to profitability, which means that I can -- without looking completely nuts -- actually factor in a breakeven-or-better mobile group in a longer-term projection of Intel's financials.

What if Intel doesn't deliver?
If Intel doesn't deliver in 2016, then I am going to start viewing Intel from the perspective of its other businesses: PC, data center, and Internet of Things. Now, I have a lot of confidence that Intel can really do well in the data center, and its Internet of Things business is small, but highly profitable and growing quickly.

It's the PC market that could make-or-break the investment case for me.

Intel's management has signaled that over the long-term, it can grow PC sales by "low single digits." If that's true, then -- coupled with the strength of all of its other businesses -- Intel still looks like it could have a lot of upside. Low single digits off of such a huge revenue and profit baseline would actually be solid.

Any smartphone or tablet growth in addition to a stable and healthy PC business could mean significant gains for Intel stock.

The risk, though, is that if the PC market winds up being far worse than expected, then mobile (tablets and phones) goes from a nice incremental opportunity to a necessary part of keeping Intel on a growth path.

Although one might reasonably argue that having a strong presence in tablets would be enough to offset any PC weakness, the average selling prices of tablet chips is significantly lower than that of PC chips. Further, in something like a smartphone, Intel can potentially capture more chip value than it can in a tablet.

And, finally, there are simply far more smartphones sold than there are tablets sold.

Why "one more shot"?
The final point to address is, why am I giving Intel only one more shot? The fact of the matter is that the mobile incumbents are growing stronger with each passing year. Qualcomm (QCOM 4.74%) and MediaTek keep growing their mobile chip businesses each year, which gives them more ability to build new, more competitive products.

At some point, there is a "point of no return" where the incumbents are so far ahead technologically and from the perspective of the business relationship required to succeed, that the only way in would be to outright buy one of those incumbents.

Buying Qualcomm's chip division is almost certainly out of the question for Intel, and MediaTek would be a tough one, too, given that its market capitalization is about $26 billion as of writing (and Intel would need to pay a premium to that).

I think that if Intel can't deliver with ground-up, segment-targeted parts built on its 14-nanometer manufacturing in 2016, then, frankly, it becomes hard to believe that the company's mobile group will ever succeed.