Can we moderate our overdependence on Asia when it comes to the microprocessors used in everything from the appliances in our homes to the laptops we use and the cars we drive?
In early September, the Commerce Department unveiled its implementation plan to disperse $50 billion from the CHIPS Act for subsidies to build U.S. chip factories and support U.S. chip research and development. Just this month, new restrictions were placed on China’s ability to buy and manufacture certain high-end chips used in military applications. The export controls also affect US companies that export semiconductor manufacturing equipment to China.
Companies are already announcing investments to reduce U.S. dependence on Asia for semiconductors. Intel Corp. plans to spend $20 billion on a new manufacturing plant in New Albany, Ohio, which is expected to be operational by 2025 and be one of the largest silicon production sites in the world. Taiwan Semiconductor (TSMC) and Samsung (which has already announced plans for a $17 billion chip plant in Texas to open in 2024) are also committed to bringing chip manufacturing back to American shores.
At the same time, investments in artificial intelligence (AI) and machine learning (ML) in the semiconductor industry are growing to increase efficiency in ways we never imagined. In today’s era of extreme automation, AI, along with digital twin technology, has the capacity to speed up the chip design and manufacturing process and in turn help us bridge the supply-demand gap more quickly.
Digital twins—virtual representations that serve as real-time digital counterparts of physical objects or processes—have evolved significantly since their first practical application at NASA in 2010 to improve simulation of the physical model of spacecraft.
Today’s digital dual technology allows chipmakers to improve performance while operating at full capacity without any interruptions. Companies such as LAM Research, Bosch (which uses a digital twin in one of its German semiconductor factories), and Applied Materials (a leader in engineering solutions for materials used to make nearly every new chip and advanced display in the world) already use surrogate models to machine learning that are more accurate and up to a million times faster than traditional physics-based simulations.
Technology startups such as Tignis (one of our portfolio companies), AspenTech and Ansys are already pioneering advances using digital twins to optimize industrial operations and make AI and ML accessible to almost every application.
With AI poised to play a key role in process control and process modeling and available for use in every area of engineering simulation, there will be a huge opportunity to disrupt the manufacturing industry by delivering significant improvements in yield, quality and productivity.
Digital twin modeling can therefore prove invaluable to the chip manufacturing process, contributing to a more streamlined design and manufacturing process while reducing reliance on physical prototypes.
However, while some chipmakers already use digital twins to create development models, the technology is not widely used to optimize production. This is surprising when you consider that by using data already available, digital twins have the capacity to help chipmakers better determine whether production targets are adequate and, if not, what even a modest increase in production might mean.
By recreating what a physical system looks like in the cloud, manufacturers can identify key learnings and achieve even greater increases in capacity—all without the risks associated with traditional methods.
With private and public funds at stake, digital twins could be a game-changer for chipmakers, manufacturers and consumers.
Chris Rust is the founder and general partner of Clear Ventures.
The opinions expressed in Fortune.com comments are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Condition.
More required comments posted by Condition:
This story was originally featured on Fortune.com
More from Fortune:
I proudly wake up at 8:59am, one minute before I start my remote job. There are thousands like me and we don’t care what you think
You may have Crohn’s disease, rheumatoid arthritis or lupus because your ancestor survived the Black Death
The stunning decline in housing in one chart: Prices have fallen in 51 of these 60 cities, and there’s more to fall
Let’s not go back on it: these 10 corporate buzzwords are the most hated in America