Samsung announced the 2nm plan to compete with TSMC Intel, who is the leader?
As the semiconductor manufacturing process enters the nanoscale, the competition is becoming more and more fierce, and the 2 nm process is considered to be an important breakthrough in the new generation of semiconductor technology, so major companies in the world are actively deploying. Earlier, TSMC and Intel announced that they would launch a 2nm process in 2025 to seize market advantages.
In addition to TSMC and Intel, Samsung has also recently announced its own plans for a 2nm process. According to a report by South Korea**, Samsung plans to start manufacturing the 2 nm process in South Korea by 2025 and plans to invest 50 billion won in the 2 nm process by 2047.
This means that Samsung has to face competitors like TSMC and Intel. So, who can beat these three companies?
The 2 nm process technology and the next-generation process technology are almost on the same starting line, and it is difficult to distinguish the winner from the loser. However, we can analyze this problem from three aspects.
i.In terms of comparing technical merits.
In the battle for the 2 nm process, technical strength is key. Samsung, TSMC, and Intel are leading in the semiconductor industry, each with a strong R&D team and its own patented technology.
Samsung has rich experience and technical reserves in semiconductor manufacturing, and has made great progress in process technology in recent years. Samsung has been using GAA technology since 3 nm, and in this regard, Samsung has an absolute advantage, and the 2 nm process is also essential.
As the world's largest semiconductor manufacturer, TSMC has always been at the forefront of the world in process technology. Take the 3 nm process as an example, although Samsung is the first in the world to mass-produce and use the most advanced GAA process, TSMC still uses the FinFET process because of the advantages of technology accumulation, so it has surpassed Samsung in the 3 nm process and won more consumers.
According to the news announced by TSMC, the 2 nm process will use a nanotransistor architecture and a low-power process, which can increase its efficiency by 10% to 15%, and save 25% to 30% at the same speed.
Intel also has an advantage in process technology and may be the forerunner of the 2nm process. Intel plans to release 18 nm process, ahead of TSMC and Samsung's 2 nm process.
In addition, Intel is also focusing on designing custom IP cores to improve the performance of the chip and reduce power consumption. The 18 A process also uses Ruibang FET architecture, GAA transistor, and back-end energy delivery technology. This technology is expected to reduce power consumption while increasing logic density and clock speed. There is no doubt that this is an important breakthrough in the 2 nm process.
Technically, all three models have their own advantages, but it's hard to say which is better or worse. There is only one way to say that Intel may be more and more advantageous and ahead of the curve.
ii.Comparison of production efficiency with output.
In addition to technical strength, production efficiency and performance are also important factors that affect customers' choice of products. In the commercial 2 nm process, process efficiency and yield will directly affect the company's profitability and competitiveness.
In contrast, Samsung does not have much of an advantage in terms of production efficiency and yield. As a comprehensive semiconductor company, Samsung has a complete industrial chain and first-class chain system, and has a stronger ability to control the production process and quality. However, the 5 nm process is greatly affected by the yield rate, so there are not many customers.
Despite this, Samsung continues to invest in improving production equipment and processes to improve production efficiency and yield. It has been reported that the yield rate of the company's 3 nm process has exceeded 60%, and new processes are being developed to improve the yield. If Qualcomm's 2 nm chips were manufactured by Samsung, the 2 nm process would be a huge step forward.
TSMC has extensive experience and strong technical force in production efficiency and yield. TSMC uses advanced production equipment and process technology to continuously optimize the process to ensure yield and cost.
In addition, TSMC has also established long-term cooperative relationships with a number of chip design institutions to meet customer needs and increase production capacity. For example, Apple has booked every new process two years in advance, and TSMC is constantly improving its advanced process technology.
Intel also has its own advantages when it comes to production. The company attaches great importance to internal production and research and development, and has advanced production equipment and technology. In addition, Intel also uses advanced design tools and processes to improve design efficiency and yield.
Of course, for Intel, there are relatively few such external contract services, and there is not much information about the 2 nm process yield issues. At present, only Kissinger has publicly claimed that Intel's 18 A process technology is ahead of TSMC's N2 process and cannot be surpassed within two years.
In short, as far as the available information goes, TSMC has"Preemptive strike"benefits. The other two, however, will not necessarily lose.
iii.Research on market demand and chain management.
With the popularization of technologies such as 5G and the Internet of Things, the demand for high-performance, low-power chips will continue to increase. Therefore, how to better adapt to market demand and the requirements for the management of the first chain has become the key to winning in the competition.
Samsung, TSMC, and Intel are all aware of the needs of this market and are actively looking for opportunities to cooperate with related enterprises. Samsung has established partnerships with many chip design companies, TSMC has close ties with many of the world's chip design giants, and Intel has focused on working with PC and server manufacturers.
TSMC has rich experience and strength in chain management. It has established long-term partnerships with many first-class merchants from all over the world, which can well coordinate the first-class chain resources and reduce costs. In addition, Samsung and Intel also have rich experience and advantages in this field.
Before this comparison, TSMC had more experience and higher power consumption, while the gap between Samsung and Intel was not large.
In general, the 2 nm war has begun, in which Samsung, TSMC, and Intel will become the big three in this race. However, it is still difficult to estimate who will emerge as the winner.