Tech

Samsung Loses 3nm Chip Orders to TSMC for Google and Qualcomm

Spread the love

Samsung Electronics’ chip manufacturing division is struggling hard with building well-organized 3nm chips. This forced major customers like Google and Qualcomm to shift to Samsung’s rival, TSMC, for their next-generation processors. 

The South Korean media reported that TSMC has just secured orders from Google and Qualcomm for next-gen 3nm because Samsung Foundry is having issues with yield and power efficiency. The new report from Business Korea suggests that most firms will deal out tier orders to TSMC for next-gen 3nm chip designs. As of today, 7 huge brands are using TSMC for 3nm designs, including NVIDIA, AMD, Intel, Apple, Qualcomm, MediaTek, and Google, which are focused on looking forward to the use of TSMC’s 3nm manufacturing process.

Google is shifting to TSMC to make their next-generation Pixel phone processor because Samsung Foundry is having problems with its 3nm chip technology. Meanwhile, Samsung Foundry is facing multiple issues, such as their 3nm chips (SF3E) having a lower yield and being less power efficient than expected, and it is also struggling to catch up to TSMC’s dominance. Google will use TSMC to make their next-gen Tensor G5 processor for Pixel 10 phones, maybe because TSMC’s 3nm technology is performing better. 

Noticeably, the Korean brand was the one that announced 3nm mass production back in 2022 for the first time. Apart from this, the next-gen Exynos 2500 processor that was developed by Samsung System LSI Division and produced through its in-house Samsung Foundry 3nm process is also reported to have “disappointing yields,” reports Business Korea. 

It is reported that TSMC holds a foundry market share of around 61.7% as it stands, with Samsung Foundry dropping to just 11% in the first quarter of the year. The representative also pointed out a few issues in heat management and stated that “heat issues in semiconductors have been a longstanding challenge for leading chip manufacturers over the past 20 years, but it has become a crucial issue with the onset of the AI semiconductor era. In mobile chips, heat issues can compromise the entire structure of smartphones, and for server chips, the heat generated by one server rack can spread like wildfire, potentially causing overload across the entire server.”.

Related Articles

Back to top button