For the mass production plan of the 3nm process, TSMC officially announced that it will mass production later in the fourth quarter, and stated: "The yield rate is high and the demand from customers is also very high!" Is there really a demand for the 3nm process in the current market? Who can win the competition with Samsung?
TSMC can succeed
TSMC has always been committed to the research and development of chip process technology, and has never planned to develop chips on its own, which has reserved enough space for customers, and customers also trust it very much. However, Samsung is different. The business involved is too wide. Chip manufacturers face the risk of technology being misappropriated when orders are delivered. Without customer feedback support, Samsung's technology has always fallen into the trap.
Samsung is the world's top memory chip manufacturer, and the model it adopts is self-developed and produced. Therefore, this model is also wanted to adopt on processor chips, but the self-developed Orion chip has reached a point where no one cares about it, so we can only focus on the foundry field.
As Zhang Zhongmou said before: "Taipei's years of hard work cannot be compensated with money!" , which is rich and powerful, has well verified this view. Although the R&D funding investment is on par with TSMC in terms of technology, it is far from the yield rate and performance performance. This is also the reason why Qualcomm is unwilling to continue delivering orders.
But TSMC did make a frustration in the 3nm process this time. As the chief supporter of advanced processes, Apple gave up the use of the first 3nm N3 process due to cost issues. focuses more on mass production of the N3E process next year. TSMC has already given up mass production of the 3nm process this year, but it seems that the plan has changed at present.
Although the N3 process cost is too high, TSMC has not claimed to give up completely. It is just rumored that Apple and Intel are not satisfied with the process, but based on recent statements, they are still quite confident. They once again stated that the 3nm process meets expectations and will usher in mass production in the fourth quarter.
and specifically emphasizes "high yield rate"! It is to reflect its advantages. Since Samsung adopts the latest GAA process, many technical problems have not been solved yet. It can only reach 10%-20% in yield rate. However, TSMC adopts mature old technology, and the yield rate will definitely reach more than 50%.
The higher the yield rate, in addition to making customers more at ease, it can also better control the cost. The word "late" in TSMC's statement has a certain meaning, which may mean that the mass production time will be at the end of 2022, but it may take until 2023 to truly produce OEM for customers. Does Samsung still have a chance? Does Samsung still have a chance to overtake
?
TSMC's yield rate exceeds 50%, while Samsung is only less than 20%. Chip manufacturers naturally know how to choose. At the same time, TSMC also has a very big advantage in the number of EUV lithography machines. Although Samsung has been urging ASML to be delivered as soon as possible, due to the current global situation, there are problems in the supply of raw materials, and the production capacity is also greatly reduced.
Therefore, in the next five years, Samsung does not have much hope to catch up with TSMC. The current demand for advanced processes is declining. Therefore, the competition between Samsung and TSMC is actually for market position. Samsung knows that it is behind in technology, so it will spare no effort to strive for the lead in technology.
But if Samsung is persistent in improving technology and does not focus on improving quality and building production capacity, even if it can achieve a process below 3nm, it will be futile and will not gain the trust and support of customers at all.
Currently, both companies have officially announced plans to realize the 2nm process in 2025. Samsung even claimed that it would realize the 1.4nm process in 2027, but given the current market demand, the 5nm process has already experienced overperformance, and there will be no demand in this area in the short term.
According to Samsung's development trajectory, it is very difficult to achieve profits in the future. The 5nm Snapdragon 888 order delivered by Qualcomm previously had problems such as severe heat generation and high power consumption. After the order was returned to TSMC, these problems were solved accordingly. Customer distrust has brought "crisis" to Samsung.
If Samsung does not pay attention to these issues, no matter how much money it has, it cannot change its market position. Relying solely on the top of the process is ineffective, and the final decision is still on the customer. What do you think about this?
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-