TSMC's 2nm process is reported to be experiencing large-scale adoption from chip manufacturers, with the likes of NVIDIA, AMD, and ASIC designers battling for capacity.
TSMC's 2nm Process Capacity Is "Heavily Contested" Between Mobile & HPC Customers
With the AI frenzy, we have seen HPC customers accounting for a significant share of TSMC's total revenue, and it appears that, with N2-class nodes, the influence of chip manufacturers will only increase. Based on a report by Ctee, it is discussed that 2nm customers are reporting the biggest-ever capacity, since mobile clients like Apple and Qualcomm will drive the initial adoption, but following them, the spotlight will shift to AI giants. And then the more difficult part will probably be how TSMC manages production yields, given how massive chip packages have become.
As far as 2nm adoption is concerned, it is anticipated that AMD will use the process node for its Instinct MI400 series, slated for H2, while NVIDIA's next major upgrade will come with Feynman, featuring the A16 (1.6nm). At the same time, ASIC players like Amazon and Google will also take up a significant share of TSMC's N2 production for next-gen architectures, indicating that the race to source high-end processes has ramped up significantly over the years.
For those unaware, with each process upgrade, i.e from 5nm to 4nm and 3nm, TSMC is seeing records breaking when it comes to customer adoption, mainly since leveraging Moore's Law has become important more than ever in recent times, where compute capabiliites have taken a priority over all other aspects. And, given that TSMC is currently leading the foundry segment, all of the cumulative demand arrives at the Taiwan giant's door, as there are few alternatives out there for foundry plays.
NVIDIA's CEO had recently been talking about how TSMC would need to ramp up production capacity to keep up with the AI frenzy, going so far as to say it would "doubling production" in a decade. With the world's largest infrastructure buildout, there's too much demand for TSMC to keep up with, especially when you factor in OSAT packaging services as well.
About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.
Follow Wccftech on Google to get more of our news coverage in your feeds.
