The 3nm GAA process might have beaten TSMC’s 3nm ‘N3B’ variant in the launch timeline, but it will matter little if Samsung cannot increase those yields where it makes sense for clients to place orders with the Korean giant. A rumor claims that the foundry had a respectable yield target for both its first and second-generation 3nm GAA node, which was 70 percent, but it has yet to reach that threshold, which would explain why Samsung has failed to attract new customers to this technology. In fact, one rumor claims that second-generation 3nm GAA yields are abysmal, where they have not even reached one-third of the intended target.
On yeux1122’s blog, Samsung’s current yields for both its first and second-generation 3nm GAA process have been revealed through domestic articles and it is not a pretty picture. The company has obtained some respectable figures for the first iteration of its 3nm technology, which is also called ‘SF3E-3GAE’ and it is between 50-60 percent. While this number is closer to the initial 70-percent target, Samsung still needs to reach production to that level for customers to justify placing orders on this lithography.
A major reason why the Snapdragon 8 Elite is exclusively mass produced on TSMC’s 3nm ‘N3E’ architecture is because Samsung reportedly lost out on orders from Qualcomm due to low yields. As for the second-generation 3nm process, the yields are substantially worse, standing at a terrible 20 percent, making it less than one-third of Samsung’s initial target. At this stage, it is unlikely that the company will gain any sympathizers over failing to show any reasonable progress, with even South Korean firms that previously placed orders with Samsung have shifted to TSMC’s camp for its more
Read more on wccftech.com