It's no secret that Samsung (NASDAQOTH: SSNLF) is the "other" major smartphone player next to Apple (AAPL 1.27%). With a massive marketing budget, and an R&D budget that allows the company to try many different things until it strikes gold, it has brute-forced its way into becoming almost as profitable as Apple.

Now, while Apple plays exclusively to the upper echelon of the handset market, Samsung isn't afraid to get down and dirty in the lowest-end segments, as well as try to go toe-to-toe with Apple in the highest end. Since Samsung holds 32% of the smartphone market, Intel (INTC 0.64%) -- which is looking to gain meaningful traction in the handset market -- will need to win sockets with Samsung.

Understanding the situation
Samsung has its own applications processor division and spent a good portion of its recent analyst day boasting that it would have a 64-bit applications processor next year (based on an ARM design), and then sometime later release a 64-bit CPU core of its own design (but still compatible with the ARMv8 instruction set). While the number of "bits" doesn't determine the performance and power of a chip (everybody will have 64-bit mobile processors within the next few years), the key is that Samsung is very serious about doing its own processors in-house.

Given that Samsung has its own semiconductor manufacturing plans and design teams, there is likely to be a bias toward the company's in-house chip designs. That being said, when it makes sense to do so, Samsung will use chips from Qualcomm (QCOM 1.41%). Over the last year, Qualcomm's chips have offered better performance-per-watt and tighter integration than Samsung's internal Exynos chips, which has resulted in the vast majority of flagship products (Galaxy S IV, Galaxy Note) being shipped with Qualcomm silicon rather than Samsung silicon.

So, what does Intel need to do?

Intel isn't there, yet
Intel's job is clear: it needs to build -- by far -- the best chips in the industry. Samsung, which makes most of its corporate profits from the sale of handsets and tablets, wants to sell the fastest, hottest, and wildest mobile devices on the planet. If its own internal teams can build the best silicon, then the company will use those chips.

However if, say, Intel were to build chips that offered significantly more performance at the same or lower power than Samsung's and Qualcomm's chips, then there is no reason that Intel couldn't ship chips into a future Galaxy S6 or Galaxy Note 5.

But, that's the rub -- today, Intel's smartphone designs are lacking. The firm's upcoming Merrifield system-on-chip will be a dual core chip fighting in a quad core world, and without an integrated modem, anything but a knockout processor isn't going to win the sockets at the major OEMs. The company's upcoming Moorefield processor, which should be a quad core iteration of Merrifield, should do much better, but it's still not clear whether this processor will be good enough to yank Qualcomm or Samsung out of the flagship smartphone sockets.

What about Broxton?
In mid-2015, Intel will release its next generation Broxton system-on-chip for handsets. This will be built on Intel's 14-nanometer FinFET process (against competing parts, which will likely be built on TSMC's 20nm process), plus sport the company's next-generation Goldmont CPU core and Gen 9 graphics.

It still won't have an integrated cellular baseband, but given that the company has signaled that it will have an "increased level of integration," it wouldn't be surprising to see connectivity (Wi-Fi, GPS, Bluetooth, NFC) integrated.

If this chip can deliver knockout performance per watt, then there's a good chance that Samsung could use the part in its flagship smartphones. It would need to be far ahead of the competition, but if Intel can deliver on a great architecture that fully leverages the advantages of Intel's 14-nanometer process, then it's difficult to see Intel actually failing to deliver here.

Foolish bottom line
While the smartphone market will be a much tougher nut to crack than the tablet market, what's required for Intel to win here is consistent execution on a leadership roadmap. It won't be enough for Intel to deliver a "good" solution if it is to displace both Samsung's and Qualcomm's teams. But, if Intel can knock it out of the park with Broxton, then the company could win major sockets at Samsung, driving significant upside to the its mobile business and share price.