The progress of Samsung’s 2nm GAA appears to be significantly better than its 3nm GAA, with its poor yields of 30 percent a primary reason fabless semiconductor companies flocked to TSMC to place orders. Thankfully, with the introduction of the Exynos 2600, which is the company’s first 2nm GAA chipset, the company has solidified its position as a worthy competitor in the foundry business, but as far as yields go, the latest figures from an investment firm indicate that the company needs to make significant strides if it wants to obtain its profitability target by 2027. Then again, the situation may not appear to be as bleak as the numbers suggest, as we discuss the matter below.
KeyBank issues a conservative analysis, believing that Samsung’s 2nm GAA isn’t as promising as previous reports have suggested
Previously, Samsung was said to have achieved 50 percent yields for its 2nm GAA process, with the company now diverting its focus to the second-generation iteration called SF2P, which the foundry giant was earlier reported to have completed the basic design. With this kind of progress, it is difficult to fathom that KeyBank, as reported by Korean outlet Chosun, would have a conservative outlook of Samsung’s 2nm GAA yields, which is less than 40 percent. Keep in mind that the Exynos 2600 is already being explored for foldable smartphones like the Galaxy Z Flip 8, and there’s even more positive news.
The Korean titan’s Taylor plant in the U.S. was originally being set up for 4nm production, but the company is now completing the transition to the 2nm wafer manufacturing, with ASML’s EUV machinery’s test operations expected to commence in March this year. In short, the company is being positioned for a worthy alternative to TSMC, with customers likely viewing Samsung as a lifeline because the Taiwanese semiconductor behemoth’s supply is being choked by insane demand. The situation is so dire that customers are paying up to a 100 percent premium to secure allocation.
Assuming Samsung’s 2nm GAA yields were below 40 percent, it would probably have never formed a multi-billion-dollar partnership with Tesla. However, we will point out that, when the mass production of the company’s Exynos 2600 was first reported on, the estimated yields were mentioned to be 50 percent, and this was back in September 2025. Fast forward to January 15, and the number has yet to change, which is puzzling, so either the industry hasn’t gotten a hold of Samsung’s actual progress, or KeyBank’s estimates are correct. We wish to give the benefit of the doubt, because there’s no way a foundry would land a multitude of orders without something to show for it.
We also reached out to Samsung for a comment, and our contact said that the company ‘does not comment on baseless rumors nor speculation,’ pretty much confirming that the yields figure from KeyBank is just an estimate and the actual progress is known to the Korean firm’s employees.
News Source: Chosun
Follow Wccftech on Google to get more of our news coverage in your feeds.





