The U.S. Department of Commerce’s ”Creating Helpful Incentives to Produce Semiconductors” (CHIPS) and Science Act for America Fund, a $50 billion investment to catalyze long-term growth in the domestic semiconductor industry in support of our national and economic security, was signed into law earlier this year.
Last month, the strategy for implementing this unprecedented, massive investment in technology in the U.S. was released.
And it’s not just happening there — for its part, the EU is planning its own version of a “Chips Act” to increase the production of semiconductors within Europe, while South Korea has committed $450 billion to its industry and the Japanese government is partnering with TSMC and Sony to open a new fab by EOY 2024.
While Moore’s law — the trend of semiconductor performance to double every two years or so — has shaped the industry for more than 50 years, a newer generation of semiconductor devices, technologies, and tools are emerging at a far faster pace and are touching every aspect of technology from artificial intelligence to 5G to space satellites to surgical robotics.
All Global Paths Run Through Semi
Despite the very first semiconductor devices having been created on United States soil, today the U.S. “fabricates only 10% of global chip capacity today, and provides only 3% of global packaging, assembly, and test capacity. Over two-thirds of the most advanced semiconductors currently are made in Taiwan, and since 2020 nearly 75% of new capacity for certain mature nodes has been added,” according to the U.S. CHIPS ACT
With private investment and semiconductor companies scrambling to address the issues and build inventory, the impact is still quite a ways off. The time-frame for bringing an advanced chip online is considerable, notes a VentureBeat interview with Deloitte: “The reality is, for the most advanced chips, like you would have in your computer or your phone right now, that could be a 26-week lead time from when you start the production in the wafer fabs to the point where it’s in the warehouse and shipping to the downstream customer. Then it could be another month after that to get it into a PC or a phone. You’re looking at six or seven months from when the manufacturer starts production to when it’s actually usable by a consumer.”
Not All Smooth Sailing
One area that is most at risk is the vast ecosystem of low-end chips that the industry, including advanced chip development, relies on. Building next-generation semiconductor foundries will help mitigate modern day issues around producing AI chips, server cores, and mobile processors -– but there are many other more basic semiconductor technologies like logic chips and transistors that will be impacted with this ramp-up of technology affecting the entire tech ecosystem, including generic or COTS-based components and standard parts from low-end semiconductor manufacturers.
Last week Bloomberg reported that Taiwan Semiconductor Manufacturing Co.’s chairman C. C. Wei warned that a “shortage of chips costing anywhere from 50 cents to 10 dollars is slowing down swathes of the $600 billion semiconductor industry” and alerted, without details, that a “50-cent radio chip has been holding up production of $50,000 cars.” This isn’t the first time that TSMC has raised the concern, either. Back in the Spring the company, which controls 28% of global semiconductor manufacturing, had identified “shortages across seven chip categories, and that four of them will continue to be affected throughout 2022.”
Semiconductor Industry Ramp-up Signals Mixed
Despite lengthy debates regarding best strategies to fuel implementation and innovation, efforts to ramp-up production are being announced in tandem with government investments.
Recent indications of market growth have been mostly mixed as the industry shifts amidst the demand. On one hand Korean chip-makers as a whole recorded “the first fall in factory shipments in almost three years in July.” At the same time it is estimated that “more than 100,000 vehicles are expected to be dropped from this week’s assembly lines in North America alone. Globally, the total number of nixed vehicle assemblies sits around 180,000.”
Analysts have weighed in as well. Deloitte hailed the end of the shortage as near in July, pointing to increased capacity and production (see graph below)
Global Semiconductor Capacity Shares by Region
(click image to enlarge; source: Deloitte)
(click image to enlarge; source: ASML)
For its part, J.P. Morgan revised its outlook on recovery, suggesting that a full recovery may not arrive until 2023. In particular, the firm noted that a recovery of automotive chips remains a challenge, given that fact that cars require different types of chips.
Private companies continue to step up to the demand. For example Reuters recently reported that SK Hynix Inc., the world’s second-largest memory chip maker, plans to invest 15 trillion won ($10.9 billion) over the next five years to build a new chip plant in South Korea. Additionally Samsung began producing chips at its first of three new manufacturing facilities slated for South Korea, U.S., and the just-fired up China factories. Likewise Intel is building two new factories at its existing Arizona site and breaking ground on another two in Ohio, plus an additional one in Germany.
And it’s not just the well known brands that are doubling down on investments. Micron, one of the world’s largest semiconductor companies and currently the only U.S.-based manufacturer of memory, reportedly plans to invest $15 billion through the end of the decade to construct a new fab for “leading-edge memory manufacturing in Boise.”
And all those chips fuel innovation — and 20’s Tech!
[ vws = ]