Taiwan Semiconductor Manufacturing Corporation (TSMC) has started trial production of its next-generation 3-nanometer semiconductor nodes. The report comes courtesy of Taiwanese publication Digitimes, which also reports that executives from US chip giant Intel Corporation are set to visit Taiwan later this month to finalize orders for the 3nm process. TSMC’s manufacturing technologies often have limited capacity during the early stages of their lifecycle, which sees few companies able to procure manufactured products.
TSMC’s initial production capacity at 3nm will be limited to 40,000 wafers per month
The report on 3nm trial production comes as TSMC and the semiconductor manufacturing arm of Samsung Foundry Group converge when it comes to estimated timelines for mass production of next-generation semiconductor manufacturing technologies. While Samsung expects the 3nm process to be ready during the first half of next year, TSMC CEO Dr CC Wei has made it clear that his company will enter mass production during the second half of 2022.
The launch of 3 nm trial production from TSMC aligns with estimates by Dr. Wei that were presented earlier this year in a call with investors. His company markets its manufacturing technology as the most advanced to date and believes it will be able to reap a profit from it for years to come.
As transistor sizes continue to shrink, chip manufacturers are taking years to develop new technologies, even as TSMC says it is on track to double its process technology performance every two years.
Moreover, as indicated by Tweet embed On Twitter, Intel is keen to make sure it is first in line when it comes to purchasing TSMC 3nm products. In general, the fresh silicon from the Taiwanese company’s manufacturing lines is believed to go to tech giant Apple Cupertino, which then uses it to maintain its technological advantages in smartphones and laptops.
According to chiakokhua, who also quotes Digitimes:
Intel Xix will visit Taiwan in mid-December to:
– Complete the scope of cooperation with TSMC.
– Request/guarantee that N3 capacity is not affected by Apple.
– Initiation of discussions about cooperation for N2.
– First wave of N3 with <60K power, will reach 40K+wpm in 1H'23.
And rumors of companies asking TSMC for the top spots in their latest chip technology have also surfaced before, with Advanced Micro Devices, Inc (AMD) and a semiconductor designer in Santa Clara and Qualcomm Incorporated in San Diego rumored to favor the 3nm process. From Samsung over its Taiwanese counterpart. .
Even more interesting is Digitimes’ belief that Intel will also discuss collaborating with TSMC for the 2nm process, as California’s Santa Clara also has plans for its 2nm contract. The technology, dubbed 20A (Angstrom), may make its way to production in 2024, according to Intel’s roadmap unveiled in July of this year.
This could be a full year before the TSMC 2nm node, which could enter production in 2025 if the back of the estimates given by Dr Wei turns out to be accurate.
Intel has been reported to have collaborated with TSMC for a 3nm process since January 2021, and the latest report indicates that the pair will work together on a graphics processing unit (GPU) and three data center central processing units (CPUs). Both are important products for Intel, which has not established itself in the fast-growing GPU market and competes with AMD in the data center segment, with the latter consistently showing strong sequential and annual revenue growth in the segment.