Taiwan Semiconductor Manufacturing Company (TSMC) has begun pilot production of its third-generation 3-nanometer semiconductor manufacturing node. This information is provided by the Taiwanese daily Digitimes, which also states that officials from the United States’ chip giant Intel Corporation would visit Taiwan later this month to confirm orders for the 3nm process.
TSMC’s manufacturing methods frequently have restricted capacity in the early stages of their lifespan, allowing only a few enterprises to acquire goods created by them.
The 3nm pilot production announcement comes as TSMC and Korean chaebol Samsung Group’s semiconductor manufacturing subsidiary Samsung Foundry are neck and neck in terms of expected mass production schedules for next-generation semiconductor manufacturing technology. While Samsung anticipates that their 3nm process would be ready in the first half of next year, TSMC CEO Dr. C.C. Wei has stated that his business will begin mass production in H2 2022.
Dr. Wei’s expectations from an investor call earlier this year were met by TSMC’s 3nm pilot production launch. His business is touting the manufacturing technique as its most sophisticated to date, and he believes it will pay benefits for years to come.
With transistor sizes shrinking, chip fabricators are taking years to create new technologies, despite TSMC’s claim that it is on track to double its process technology performance every two years.
Furthermore, as stated on Twitter by @chiakokhua, Intel is keen to be the first in line to obtain TSMC’s 3nm goods. In general, the freshest silicon from the Taiwanese company’s manufacturing lines is supposed to go to Cupertino tech behemoth Apple, Inc, which subsequently utilizes it to preserve its technical advantages in smartphones and laptops.
According to @chiakokhua, who also quotes Digitimes:
Intel execs to visit Taiwan mid-Dec to:
– Finalize scope of cooperation with TSMC.
– Solicit/ensure N3 capacity will not be affected by Apple.
– Initiate discussions about cooperation for N2.
– First wave of N3 capacity <60K, will only reach 40K+ wpm in 1H’23.
Rumours of companies approaching TSMC for top positions in its latest chip technologies have previously surfaced, with Santa Clara semiconductor designer Advanced Micro Devices, Inc (AMD) and San Diego firm Qualcomm Incorporated both rumoured to prefer Samsung’s 3nm process over its Taiwanese counterpart.
More importantly, Digitimes’ expectation that Intel will also consider collaboration with TSMC for the 2nm process is telling, given that the Santa Clara, California-based business is also planning its own 2nm node. According to Intel’s roadmap, which was announced in July of this year, this technology, dubbed 20A (angstrom), might find its way to the manufacturing line in 2024.
This might be a full year ahead of TSMC’s 2nm node, which is expected to go into production in 2025 if Dr. Wei’s estimations are correct.
Intel’s collaboration with TSMC for the 3nm technology has been reported since January 2021, with the most recent rumour indicating that the two will collaborate on a graphics processing unit (GPU) and three datacenter central processing units (CPUs).
Both are critical items for Intel, which is still trying to establish itself in the quickly expanding GPU industry and is battling with AMD in the data centre space, where the latter has continuously demonstrated excellent sequential and yearly sales growth.
- Sapphire’s GOPRO X080 will be an AMD Navi 22 GPU carrying 10 GB RAM GDDR6 RAM clocked at 16 Gbps
- Intel’s Core i9-12900KF tops the benchmarking charts of both the PassMark CPU Mark and UserBenchmark