TSMC Legal Conference: TSMC 2.0 and Intel IDM 2.0

When Intel entered the foundry business with IDM 2.0, I was impressed. Yes, Intel had tried the foundry business before, but this time they changed the face of the company with IDM 2.0 and went “all in,” so to speak. The progress has been impressive, and today I think Intel is ready to take non-TSMC business by offering a credible alternative to TSMC’s leading edge business. The trillion dollar question is: will Intel take TSMC’s business away from the competition? I certainly hope so, for the greater good of the semiconductor industry.

In the most recent TSMC investor call, the first since Wei took over as chairman and CEO, TSMC named its foundry strategy Foundry 2.0. This is not a change in strategy, but a new brand based on what TSMC has been doing successfully for many years, with additional products and services added to attract customers. 3D IC packaging is an obvious example, but certainly not the only one. The Foundry 2.0 brand lives up to its name and is clearly aimed at Intel IDM 2.0, which I think is interesting and a good example of Wei's sharp intellect.

I was convinced that Intel 18A would be Intel's breakthrough foundry node, but according to TSMC's investor call, that is not the case. TSMC N3 has been a huge success, with 100% of major designs winning. Even Intel is using TSMC N3. I haven't seen anything like this since TSMC 28nm, which was the only viable 28nm HKMG node. History is repeating itself with N3 due to delays in 3nm alternatives. This makes the TSMC ecosystem the strongest I have ever seen, with N3 dominating and TSMC's packaging success expanding rapidly. I would have thought that some customers would stick with N3 until the second generation N2 comes out, but I was wrong. On yesterday's investor call:

Wei Zhejia: We expect the number of tape-outs for 2nm technology in the first two years to be higher than the number of tape-outs for 3nm and 5nm technology in the first two years. N2 will provide full load performance and power advantages, with a speed increase of 10 to 15 times at the same power, a power increase of 25% to 30% at the same speed, and a chip density increase of more than 15% compared to N3E.

Wei Zhejia mentioned this before, but now I can confirm it based on the chatter inside the ecosystem meeting at a recent conference: N2 design is underway and will start tapeout by the end of this year.

I really don’t think the TSMC ecosystem gets enough credit, especially after the huge success of N3, but the N2 node was a force in its own right:

Wei Zhejia: The development of N2 technology is progressing smoothly, and device performance and yield have been completed as planned or exceeded. N2 is expected to achieve mass production in 2025, and its capacity expansion curve is similar to that of N3. With the strategy of continuous improvement, we also launched N2P as an extension of the N2 series. Compared with N2, N2P has a 5% performance improvement at the same power and a 5% to 10% power improvement at the same speed. N2P will support both smartphones and HPC applications, and is scheduled to achieve mass production in the second half of 2026. We also launched A16 as our next-generation nanosheet-based technology, with Super Power Rail (SPR) as a separate product.

Wei Zhejia: TSMC's SPR is an innovative, first-class backside power supply solution, and the industry will adopt the backside power supply solution in the future to maintain gate density and device flexibility. Compared with N2P, A16 further increases the speed by 8% to 10% at the same power, or increases the power by 15% to 20% at the same speed, and increases the chip density by an additional 7% to 10%. A16 is best suited for specific HPC products with complex signal routing and dense power supply networks. Mass production is scheduled for the second half of 2026. We believe that N2, N2P, A16 and their derivatives will further expand our technological leadership and enable TSMC to seize future growth opportunities.

Congratulations to TSMC on their continued success, which is well deserved. I also congratulate the Intel foundry team for making the change, and I hope that the 14A foundry node will provide the industry with a credible alternative to TSMC. In my opinion, without Intel, and certainly Wei's leadership and response to Intel's challenge, our industry would not be so rapidly approaching the $1 trillion revenue mark. Say what you want about Nvidia, but TSMC and the foundry business are absolutely the true heroes of the semiconductor industry, as Huang Renxun publicly acknowledged.