Samsung wants to develop a real high-end SoC by 2025 that can compete with Apple. In the years 2023 and 2024, no AP, an application processor, as Samsung calls its smartphone SoCs, should appear. This could mark the beginning of the end of the Exynos.
In any case, the adaptation rate of the company's own chips had recently become steadily smaller. While it was almost 50 percent in all Samsung devices in 2018, it was only 28 percent last year. In the overall market, Exynos only has a market share of 6.6 percent anyway, Qualcomm, MediaTek & Co overpowered. The prospects are also not rosy, which is why Samsung is regrouping the division and setting up a task force that is to develop a new Galaxy chip from this summer that can compete with Apple in 2025.
On paper, with the announcement, it looks like they're going to throw almost everything they've done and start over. Therefore, the already planned Exynos for the years 2023 and 2024 fall directly victim, the resources are diverted to the new project. According to the media report, more than 1,000 people are involved, but in the end it should be significantly more, reports the Korean newspaper Donga exclusively the foundry division unclear
As a conglomerate, Samsung is fighting on several fronts. On the one hand, it wants to play at the top in the field of smartphones and the appropriate chips, on the other hand, it also wants to show off as a semiconductor manufacturer. In the end, it didn't always go well together, one slowed the other down and vice versa – but you know that in a similar way from Intel. Samsung's 4nm manufacturing was so bad to begin with that their prestige project with AMD was pretty weak. Qualcomm announced the partial switch to TSMC just a few days ago and will get a lot more power for its own chips thanks to the better production. So far, the foundry has always been at a disadvantage against TSMC, the fact that various major customers have left makes it abundantly clear.
The question that remains is what Samsung is doing with the first ATM processes. So far, it has been said that series production for 3GAE (Gate all around Early), i.e. the first process, will start this quarter, and that an Exynos will probably be the first product. 3GAP as a slightly improved version should follow quickly, depending on Samsung's schedule here, it could produce another product at the end of 2022. In almost all new processes, Samsung is always the first customer of Samsung Foundry.
And then? Samsung's next production step is 2 nm. And so far it is planned as 2GAP (2 nm Gate-All-Around Plus) for the end of 2025. The newly targeted “Galaxy processor” for the year 2025 is likely to be aimed directly at this production stage. By then, Samsung should at least have gained a lot of experience with ATMs, TSMC and Intel will only try it from this point in time. It remains to be seen whether this will happen in the end.