Why does Samsung use two SoCs on the same models?
To be really honest, Samsung hasn’t really revealed the exact reasons behind this. However, there are several reasons that range from the very obvious to somewhat speculatory. Let’s explore a few of those.
One of the reasons Samsung chose this path, at least during the early days of Android phones, was owing to the fact that Samsung’s own Exynos chipsets featured modems that offered wider support for GSM and LTE networks used in European and Asian countries. Qualcomm SoCs, on the other hand, came with modems that offered a wide range of bands while also supporting CDMA networks in the US (which have since then been retired). It was easier for Samsung to use two different SoCs that offered near-similar performance and offered better network support for specific markets.
Makes sense, right?
The other reason is sheer economics.
Even though Samsung ‘buys’ its Exynos SoCs from its semiconductor subsidiary, getting an SoC in-house is always cheaper than buying it from a third party. Additionally, Qualcomm (to this day) has somewhat of a monopoly in the high-end SoC segment which means it can use its position to leverage for higher prices. This is one of the reasons that even though the GSM/CDMA issue doesn’t exist anymore, Samsung continues its policy of having two variants of the same phones with different SoCs. It makes more economic sense for them and lets Samsung eke out higher margins by selling Exynos variants of its flagship phones.
Then there is another argument pertaining to the supply chain. Samsung is among the largest manufacturers of smartphones in the world by volume, and making chips to satisfy Samsung’s demands could easily be the sole focus of a chip-maker. Likewise, dependency plays a part too: if there are problems at Qualcomm, Samsung can step in and equip a market with Exynos chips.
The split also allows Samsung to attack the market on two fronts and establish what works best and why. Not to mention that there is reportedly an existing contract between Samsung and Qualcomm that both sides are still committed to. In any case, it is unlikely that Samsung could produce enough Exynos chips for all of their devices, even if they weren’t tied to Qualcomm.
Now that you have sort of understand why Samsung continues with this policy of using different SoCs on its same flagship models let us now come to the burning question.
Which of these two SoCs offer better performance on Samsung flagships
Snapdragon vs Exynos: Which is better?
Many of you might be already aware of the fact that for the past few years, flagship Snapdragon SoCs were miles ahead of their Exynos counterparts in almost every aspect. The difference was so stark in 2020 that frustrated Samsung Exynos users issued an ultimatum to Samsung or face a boycott. Some people even set up online petitions to coerce Samsung to start using Qualcomm SoCs on this year’s flagship phones.
As things stand today, Qualcomm Snapdragon 888 toting Samsung Galaxy S21s still offer better overall performance compared to its Exynos 2100. However, Samsung has certainly managed to close the massive gap in performance between Snapdragon and Exynos SoCs from the yesteryears. Perhaps the ultimatum issued by Samsung Exynos users helped. Now, the reason for the massive improvement in performance with the new Exynos 2100 has been explained in somewhat detail by me in this article. Here’s a short explanation anyway.
Essentially, what you need to know is that Samsung, until last year, used its own custom CPU cores codenamed Mongoose on its flagship SoCs. While this is based on the ARM architecture, the Mongoose core has traditionally lagged behind ARM’s A7X cores in terms of performance. Starting 2021, however, Samsung decided to abandon the Mongoose core and joined the ARM CxC project (for its flagship SoCs), of which Qualcomm too is a beneficiary.
The first performance-oriented CPU to come out of the CxC project is known as the ARM Cortex X1, and both the Snapdragon 888 and the Exynos 2100 use it (with slight customizations). Effectively, the current generation of Exynos and Qualcomm flagship processors have several similarities since they are based on identical microarchitecture. This has resulted in Exynos managing to somewhat catch up with Qualcomm in terms of CPU performance.
However, one area where Samsung continues to lag behind even with the Exynos 2100 is in the graphics department. This is because Exynos uses ARMs Mali G78 GPU for graphics, while Qualcomm uses its own Adreno 660, which is known for its much better performance. But Samsung does have one ace up its sleeve to counter this.
In 2019, Samsung announced a partnership with AMD to bring mobile GPUs to Samsung Exynos chips. In 2021, just before the launch of the Galaxy S21, Samsung announced that the first GPU to come out of the Samsung AMD collaboration is under development. And if leaked performance benchmarks are anything to go by, it might turn out to be a performance monster that could obliterate not just Qualcomm’s Adreno GPU but even Apple’s all-powerful custom GPU that it uses on the A14 Bionic SoC.
While it is too early to say, there is a good chance that Samsung’s next-generation flagship SoCs could match and perhaps even leap ahead of the competing Snapdragon offering in terms of sheer performance.
If this does indeed happen, it would be a massive turnaround for Samsung. It would also be great news for millions of Exynos users who have been forced to put up with an objectively inferior product for several years while knowing that the same company sells an identical device with a much better SoC in another part of the world. And just maybe, we are looking at a future where consumers will not be worried about choosing between Exynos or Snapdragon SoCs on Samsung flagships.
This article was updated in April 2021. Old comments have been retained.