Free tools Windows power users keep installed
One-click scans. No signup required.
In a 2021 interview, then-GlobalFoundries CEO Tom Caulfield used “the other 70%” to describe semiconductor-market revenue outside what he characterized as compute-centric chips. He argued that chips for pervasive computing—such as radio-frequency, embedded-memory, mixed-signal, and high- or low-voltage applications—deserved more manufacturing investment, even when they did not need the smallest process nodes. The figure was his strategic framing, not an independently verified or timeless measure of the market.
What did Caulfield mean by “the other 70%”?
Caulfield’s argument began with an approximate split: he said compute-centric chips represented about 30% of semiconductor-market revenue, leaving the “other 70%” for a broader mix of chips. The 70% was the counterpart to his estimate, not a separate measurement established by the interview. In his April 15, 2021 EE Times interview, he called this broader activity “pervasive computing”—electronics functions spread across many kinds of devices, rather than only high-performance computing.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Regional Renaissance: How New York’s Capital Region Became a Nanotechnology Powerhouse... | $139.99 | Buy on Amazon |
He named high- and low-voltage integrated circuits, embedded memories, radio-frequency (RF) circuitry, and mixed-signal circuits as examples. These are different functions, not one technical category; the common point in his argument was that many do not depend on leading-edge transistor scaling. He said that for much of this chip mix, the benefits from continued shrinking had already been exhausted.
Why did he argue these chips deserved attention?
Caulfield’s point was about manufacturing priorities as well as market size. If important functions can be made without the smallest process nodes, he argued, investment should also support the specialized processes and ecosystem capabilities they need. In the interview, he contrasted investment in process and equipment with work on process-design kits and partner intellectual property, which help customers design for a foundry’s manufacturing processes.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problems#1 Best Overall
His case was that pervasive-computing chips were essential even if they were not the headline-grabbing compute processors attracting much of the industry’s focus. That is a strategic argument about where capacity and ecosystem investment should go—not a claim that every chip in the broader segment uses the same process or that leading-edge manufacturing is unimportant.
How did Caulfield place GlobalFoundries among competitors?
Caulfield presented GlobalFoundries (GF) as a specialist in the broader set of technologies. He said Intel’s announced foundry ambitions would make it a force in the compute-centric 30%, while GF would be a force in both that area and the “other 70%.” He also acknowledged that TSMC played across the entire market, saying it “plays in 100% of the market.”
Those were Caulfield’s characterizations in 2021, not an independent assessment of current market boundaries or each company’s capabilities today. His comparison is best understood through three strategic distinctions:
- Chip function: compute-centric processing versus pervasive functions such as RF, embedded memory, mixed-signal, and high- or low-voltage circuits.
- Manufacturing needs: leading-edge scaling versus specialized process technologies that may not require the smallest nodes.
- Ecosystem investment: process and equipment alongside design tools such as process-design kits and partner intellectual property.
These distinctions explain the logic of his pitch; they are not a complete taxonomy of the semiconductor industry.
What was happening when he made the case?
The interview appeared amid pandemic-era demand growth, semiconductor shortages, and political tension between the United States and China. Caulfield said the pandemic had accelerated demand for pervasive computing. He also described GF reallocating capacity to help keep automakers supplied. These points are his account of the company’s response and the market conditions at the time, rather than a current description of supply or demand.
Caulfield also connected the argument to debate over US semiconductor-manufacturing policy. He told EE Times, “We want to do our part, and we think it’s an important part.” The remark expressed GF’s case for a place in US manufacturing efforts; it does not establish the details or merits of any later policy proposal.
What does the “70%” tell readers now?
It tells readers how Caulfield framed GF’s strategic position in 2021: much of the semiconductor opportunity, in his view, lay beyond compute-centric processors, in chips enabling pervasive functions. It does not show that 70% remains the current share of semiconductor revenue, or that a single source has independently validated that proportion. Treat the figure as an attributed argument from that interview, not as a present-day market statistic.
Caulfield’s title has also changed. GlobalFoundries’ leadership page says he served as CEO from 2018 and was elected Executive Chairman in April 2025. In an October 2025 company post, he discussed policy tools intended to create demand for US-made chips and said the pace of GF’s proposed expansion depended on customers committing to bring more supply back home. That later policy discussion provides context, but it does not retroactively verify the 2021 “70%” framing. GlobalFoundries’ October 2025 post describes that later position.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
