Qualcomm (QCOM 1.62%) had a pretty great gig supplying high-end mobile applications processors to smartphone giant Samsung (NASDAQOTH: SSNLF). Although Samsung would routinely use its own applications processors in variants of its smartphone flagships in some regions, Qualcomm was able to get its high-end chips inside of Samsung flagships sold in many important regions such as North America.
However, during this product cycle, Qualcomm found itself essentially out of this year's Samsung flagships; the South Korean smartphone giant instead opted to use its own Exynos 7420 applications processor paired with its own Exynos modem in most regions, only pairing said Exynos chip with Qualcomm modems in regions and/or on carriers that require support for CDMA EV-DO.
Although there is optimism that Qualcomm might be able to win back some of this lost business at Samsung with its next generation Snapdragon 820 processor (from what Qualcomm has disclosed, it looks quite good), I'm not convinced that Qualcomm will definitely hold on to its Samsung business over the long term even if it wins a spot in the upcoming Galaxy S7.
Samsung is seemingly aggressively investing in its own mobile chips
Samsung has been putting out its own mobile processors for a while sold under the Exynos branding. Generally these chips have been competent, and as in the case of the latest Exynos 7420, have had the advantage of being first to utilize Samsung's leading-edge manufacturing technology, but from an architectural perspective they've not been anything all that special.
In fact, I would argue that from an intellectual property/capability perspective, Qualcomm's high-end Snapdragon chips have been much more interesting.
Although I suspect that Qualcomm's future chips will continue to be interesting for quite some time, it would appear that Samsung is investing heavily in trying to build more interesting mobile-focused chip designs.
For example, Samsung is said to be including custom-designed CPU cores in its next generation mobile processor, reportedly called the Exynos M1. It is also clear from various LinkedIn profiles of Samsung engineers that the company is working on custom mobile graphics processing unit designs.
Additionally, Samsung was able to successfully deploy its own Exynos Modem in many versions of its latest Galaxy S6/Note 5 flagships, eschewing Qualcomm's (arguably superior) modems in most models.
Finally, Samsung has even managed to sell its Exynos processors to third-parties; the Meizu Pro 5 smartphone, for example, uses Samsung's Exynos 7420 chip.
If Samsung is investing this aggressively, will it eventually completely cut Qualcomm out?
Qualcomm already suffered a heavy blow when it was designed out of Samsung's flagships this year, and although I do think Qualcomm has a good chance of winning back high-end Samsung business next year, I have to wonder if Samsung will eventually cut Qualcomm completely out of its smartphones.
Now, one could reasonably argue that Samsung might agree to use Qualcomm's chips in exchange for Qualcomm moving much of its chip manufacturing volume over to Samsung (boosting Samsung's foundry business). Qualcomm, in this case, would benefit from having a large buyer of chips and Samsung's foundry would benefit from being able to ship much of the world's mobile chip volume.
However, I think that there is a risk that when/if Samsung has a more complete and competitive mobile chip portfolio, not only could it begin replacing Qualcomm in more of its smartphones, but it could start competing with Qualcomm for chip spots in other companies' phones.
Qualcomm needs to stay a step ahead to protect its business
At the end of the day, I believe that Samsung could ultimately prove to be Qualcomm's most powerful competitor in the mobile chip business. Samsung's mobile chip portfolio isn't really there today (MediaTek is proving an increasingly powerful competitor to Qualcomm), but with the proper investment it could be.
In order to stave off this threat, Qualcomm's chip teams are going to need to stay at least a step ahead of what Samsung's teams can do.
At some point, I believe that Samsung will either significantly scale back its mobile chip efforts (in the case where it simply can't compete with Qualcomm/MediaTek) which would be a huge positive for Qualcomm, or it will actually become a force to be reckoned with in the mobile chip market, which would be an equally huge negative for Qualcomm.
