After numerous leaks and endless rumors, Samsung’s newest Galaxy S23 series has now been officially unveiled, confirming several of those leaks we stumbled upon in the past couple of months. One of those leaks that particularly caught the eye was Samsung ditching its in-house Exynos SoC, going all-in on Qualcomm’s latest offering, bringing parity to its customers across all the different markets we all craved for so long. Not only that, the Galaxy S23 series will feature a special beefed-up version of the Snapdragon 8 Gen 2 SoC, featuring the “For Galaxy” moniker. At first, this might sound spectacular, especially for those in the EU and MENA regions contemplating the idea of upgrading to the newest flagship from Samsung this year. However, I’d like to argue that ditching Exynos altogether might not be the best news to hear in the long run.
Indeed, Samsung’s Exynos SoC has come under heavy scrutiny, and rightly so for its shortcomings in both performance and efficiency when compared to its Qualcomm counterpart. This disparity reached its most blatant peak in 2020 when the Galaxy S20 series was released, with the Exynos 990-powered variant coming up substantially short in almost every aspect. Even last year’s situation with the S22 series was not that different, except that fewer markets were getting scammed by Samsung’s ploy. Having said all that, and despite feeling the Exynos 990’s heat myself — pun intended, I still think that taking Exynos away from the equation is not the best thing to happen for the industry.
It might sound a bit like an overused cliche at this point, but it still remains an undeniable truth. Competition is always good for the customers, and by pulling the plug on Exynos, Samsung is just withdrawing from the competition. But as they say, history is the best teacher. So let’s wind the clocks back to when Qualcomm was comfortably sitting on the throne.
2015: Knocking the dragon off its perch
Back in 2015, after years of dominance in the flagship space when 64-bit was merely an afterthought, Qualcomm was clearly caught lacking. A couple of years earlier, Apple induced one of the biggest shifts in the industry by introducing the world’s first 64-bit mobile CPU with the A7 SoC. Meanwhile, Samsung tried to steal its thunder by releasing the first big.LITTLE-based CPU in its coveted Exynos 5 Octa SoC, despite not featuring any 64-bit support of any sort. Qualcomm, on the other hand, featured neither of those, which was embarrassing, to say the least. Indeed, Qualcomm had to do something in response, incidentally rushing back to the drawing board. The plan was simple: bring 64-bit support to the table and integrate the big.LITTLE design. The execution, however, was far from simple, resulting in one of the worst editions of a Snapdragon SoC Qualcomm released.
While early reports of overheating were dismissed as nothing alarming for early testing of prototypes, Qualcomm’s fiasco was on display as soon as the phones hit the stores — the HTC One M9 was already displaying overheating warnings midway through an AnTuTu benchmark run. Adding more petrol to the fire, Samsung decided not to use the Snapdragon 810 chip for any of its variants of the S6 and S6 Edge that year for the same reason. At least, Samsung had Exynos to thank for bailing them out that year. The rest of the OEMs, unfortunately, had little to no options other than using Qualcomm’s disastrous chip and hoping for the best. It is fair to say, Qualcomm has wreaked havoc on the flagship market, with several phones falling victim to Qualcomm’s shambolic chipset.
Not that much later, Qualcomm attempted to remedy its losses by launching a much tamer variant of its SoC, the Snapdragon 808, taking away two of its big cores and toning down the GPU slightly in an attempt to increase its thermal headroom. This did improve the thermal efficiency significantly. However, at the end of the day, you are paying flagship money for a phone that is undeniably inferior to a phone like the Galaxy S6 that is priced similarly.
A year later, Qualcomm tried to claw back its position in the market with the Snapdragon 820, but again Qualcomm’s SoC was trailing behind Samsung’s counterpart, the Exynos 8890. It was not exactly a battle that was totally won or lost by either. However, Qualcomm’s position that year was still a far cry from its dominant position in the market a couple of years earlier. Only the following year Qualcomm was able to largely restore its reputation as the best on the market, though it can be argued that it was down to a mishap from Huawei’s side with its Kirin 960 SoC.
2020: How the turntables…
Fast forward to 2020, and it is almost the other way around. In typical fashion, Samsung launched its S20 series in February, powered by the Snapdragon 865 in the US, South Korea, and China, while the Exynos 990-powered variants hit the shelves in the rest of the world. On paper, there were slight differences between both chipsets, most notably the way each CPU was configured. While the Snapdragon 865’s CPU had 1 prime core and 3 big cores in a 1+3+4 setup, the Exynos 990’s CPU swapped one of the big cores for a prime one in a 2+2+4 setup, which might have looked a bit promising. In reality, though, the Snapdragon-powered variant offered a vastly superior experience compared to its Exynos-powered counterpart, leaving nearly everyone visibly envious of those lucky enough to lay their hands on a Snapdragon-powered model.
This disparity was even further exacerbated with the Note 20 series, as Samsung released its Note 20 and Note 20 Ultra phones packing an even more refined Snapdragon 865+ chip inside in the same select markets as the S20 series earlier. Meanwhile, the rest of the world got the Note 20 series in their stores with the same Exynos 990 chipset inside. While the Exynos 990 was not exactly the flamethrower the Snapdragon 810 was back in the day, its shortcomings were on display for all to see as echoed by several users and reviewers and further confirmed by several tests. Even Samsung’s most loyal fans couldn’t stand by their side as they get scammed into paying the same money for a phone that is noticeably inferior to the same phone in another region.
Even more unimpressive was Samsung’s statement on the Exynos 990’s debacle, claiming that both variants are subjected to the same testing scenarios to ensure “consistent” performance. Irony abounds…
2021: Trading blows
Still, at least Samsung ought to have learned something from this fiasco, which looked like the case the following year as the Exynos 2100 and Snapdragon 888 traded blows across the board. This was also the first year Samsung completely ditched its custom Mongoose CPU cores in favor of ARM’s off-the-shelf Cortex X1 ones. Of course, Qualcomm’s tight grasp on the crown of the GPU performance remained just as tight. But as far as the CPU is concerned, Samsung’s Exynos 2100 was neck-in-neck with its contender from Qualcomm. This came as a huge sigh of relief for a lot of people, particularly those who have been eyeing an upgrade to Samsung’s Galaxy S21 series that year.
As for those who wanted to hold on for another year, confidence in their decision was sort of justifiable based on the fact that Samsung might have fixed its formula for the CPU part at least. Even more promising was the news that the Exynos 2200 the following year would debut the first GPU product of Samsung and AMD’s lucrative partnership, the Xclipse 920 GPU. If anything, this looked like a perfect plan for Samsung to further strengthen its position as a strong contender in the SoC manufacturing space. With the new GPU, Samsung would finally address what many deemed its weakest link for many years before. Things were looking very promising for the South Korean giant, setting high expectations among its customers while looking like a real threat to Qualcomm’s dominance. Some even anticipated that with AMD’s experience and prowess, even Apple’s continued dominance would be put in jeopardy. So how did that turn out in the end?
2022: Back to square one
In short, the word underwhelming would be a massive understatement. While Samsung’s brand new GPU put out incredibly modest gains over its predecessor, sustained CPU performance issues reared their ugly head once more, so much that it would result in substantial performance penalties that were not exhibited by its Qualcomm-powered counterpart. It’s not like Qualcomm’s Snapdragon 8 Gen 1 was flying out of the gates either, as it too exhibited sustained performance issues. Yet the discrepancy was all there to see, with several users voicing their concerns with the performance of the Exynos model, citing multiple occurrences of lags and sluggishness across the UI along with other issues. Some even deemed the Exynos-powered model not usable as a daily driver. This was extremely disappointing from Samsung for its customers to pay top dollar for a supposed flagship phone, only for it to exhibit such symptoms.
https://twitter.com/AdamConwayIE/status/1496481958016262148
This brings us to this year’s flavor of Samsung’s flagship phones. A quick look at the spec sheet of the Galaxy S23 series would indicate that this is more of an iterative upgrade over its predecessor from last year. If there is anything to write home about, it would certainly be the S23 Ultra’s mammoth 200 MP main camera sensor, and the increased battery capacities for the smaller S23 and S23+ models. However, one notable omission that might have flown under most people’s radar resides a bit deeper in the spec sheet. After so many years, this is the first time Samsung decided to launch its phones with the latest and greatest chipset Qualcomm has to offer: a souped-up version of the Snapdragon 8 Gen 2. People no longer have to make do with Exynos’ deficiencies in any part of the world, or attempt to import the Snapdragon variant from elsewhere. If anything, it looks as if Samsung has finally thrown the towel, as people rejoiced their good riddance to Exynos’ shenanigans. But as we zoom out to the bigger picture, there is more to it than that.
2023: Danger averted, but at what cost?
With Exynos completely phased out, and the continued sanctions on Huawei crippling its endeavors of developing its own in-house Kirin SoC, it is practically only Qualcomm in the market manufacturing the SoCs for the majority of the flagship phones. It’s true, MediaTek surely wowed us all last year with its Dimensity 8100 and 9000 SoCs, yet only a couple of phones featured these SoCs inside, with Qualcomm taking the lion’s share despite its Snapdragon 8 Gen 1 not being the best release.
This all brings us to the old cliche once more — competition is always good for the customers. We cannot afford to lose yet another contender, surrendering nearly the whole market to Qualcomm alone. We have all seen what tends to happen when Qualcomm rests on its laurels. And if the matter of their own phones and chipsets doesn’t seem to trouble them that much, it must be noted that Samsung has another client of vast significance to be satisfied.
Now that Google is heavily relying on Samsung’s foundry to manufacture their Tensor chipsets, Samsung’s current situation surely doesn’t bode well for Google. Apart from the slightly bizarre naming, Qualcomm’s Snapdragon 8+ gen 1 most notable change was the move from Samsung’s 4nm manufacturing node to TSMC’s one. That alone was sufficient to bring up to 30% improved efficiency, addressing one of the biggest concerns with its predecessor. If Google is still going to bet big on Samsung’s foundry for its Tensor chipsets for the foreseeable future, then Samsung has got to step up its game. Failing to do so may see them losing Google as their client, which will surely be a big blow to their business.
Having said all that, despite how weird it might sound, I think that it is the best for everyone if Samsung doesn’t wield the axe in its Exynos SoC. It just has to be better, not dead.
Featured-Image: GizNext
