When it comes to semiconductors, driving isn’t everything

Take the editor: Undeniably, we usually spend a lot of time talking about the leading semiconductor manufacturing. This is a common mistake everyone makes when discussing semi-finals, and one we are as guilty of as anyone. The world rightly focuses on the dearth of companies able to operate at the forefront, but there is a lot of semi-finished business.

Editor’s note:
Guest author Jonathan Goldberg He is the founder of D2D Advisory, a cross-functional advisory firm. Jonathan develops growth strategies and alliances for companies in the mobile, networking, gaming and software industries.

We recently went looking for data on fab capacity by operation decade, and all agreed that the leading expert on the subject is Bill McClain at IC Insights. He maintains one of the strictest paradigms on the subject, and rightly charges a premium for his reporting. This is must-read material for anyone planning a multi-year semifinal roadmap.

A quick Google search yielded this snippet of IC Insight data, and it tells an important story…

More than 90% of the world’s semiconductor capacity operates at 10 nanometers or higher. We could argue about where to draw the line, but it’s safe to say that the vast majority of capacitance runs at the trailing edge.

This is important for a number of reasons.

First, when the world ran out of semiconductors in 2020/2021 — most of that shortfall was occurring in these more mature operations. TSMC’s flagship customers were all able to get most of the capacity they needed in 7nm, but there was a real pain for industrial and automotive customers.

These companies needed simple parts such as Microcontroller Units (MCUs) and Power Management ICs (PMICs), and these products are generally produced on legacy nodes. Today, even as supply shortages have turned into excess inventory in many categories, vintage products are catching up with pent-up demand from two years ago.

Second, the US government is currently struggling to decide how to allocate the $52 billion Chips law funds. if the purpose Of that money is just to bring the leading operations back to the US, and then go ahead and give all the money to Intel. They’re going to distribute $7 billion or $8 billion to shareholders and go ahead with their plan to catch up with the manufacturing that they’re going to have to do anyway.

On the other hand, if the goal is to truly secure the quasi-American supply chain, perhaps a better plan would be to split that money more broadly. Ideally, they would spend the money to plant lots of seeds that lead to the formation of a new company and in basic academic research, which can then be commercialized by the private sector. Unfortunately, there are no easy mechanisms for doing this yet, so another approach would be to divide the money among the wide range of US companies involved in semi-finished manufacturing, as long as they commit to increasing US capabilities. This means not only finished products and foundries, but also needs to include tooling companies, robotics providers, and chemical makers—the entire supply chain. Intel should get some of that money, but not most of it.

According to the Semiconductor Industry Association, the CHIPS Act has been positive side effect By incentivizing the private sector to invest about $200 billion in semiconductor production in the United States.

Finally, these numbers should remind us that the story is broader than just TSMC and Samsung. There is still a lot of important and interesting work being done in trailing edge foundries.

The clearest example of this is Global Foundries. GloFo is not at the cutting edge of the silicon industry but has carved out some very large “niches” such as silicon-on-insulator (SOI) and silicon carbide. And while they don’t have the near monopoly that TSMC and Samsung enjoy with 7nm, they do come close to many of their SOI lines. If the US loses access to TSMC for any reason, GloFo arguably will be as important a part of the solution as Intel.

Source link

Related Posts