The U.S. Congress and the Biden Administration are hurtling down the track toward investing billions of dollars in the domestic semiconductor manufacturing industry. But there is a flaw in their approach because they are ignoring the broader manufacturing infrastructure that is needed to combine semiconductors with other parts to create reliable electronic systems.
In January 2021, the U.S. Congress passed the CHIPS Act, a $53 billion effort to bring the most advanced semiconductor wafer manufacturing back to North America. The momentum is continuing with the proposed Facilitating American-Built Semiconductors (FABS) Act, which would provide a permanent 25% investment tax credit for investments in semiconductor manufacturing; and the $250 billion U.S. Innovation & Competition Act (USICA), which would boost federal R&D investment in 10 high-tech fields including high-performance computing and semiconductors.
But wafer fab is only one piece of the puzzle. Finished wafers must be placed in some form of package, such as an integrated circuit (IC) or other device, which is then assembled along with resistors, capacitors, metal traces, and other components on a printed circuit board. That circuit board is then interconnected with other systems to create a car, a computer, a cell phone, or an airplane.
Indeed, a computer chip is like the engine of a car. Without all of the car’s other parts, the engine can’t accomplish much. Likewise, without other electronic components and interconnections between them, a computer chip cannot function. The modern world runs on electronic systems, not just computer chips.
Twenty-five years ago, most original equipment manufacturers (OEMs) such as IBM, Sony, Ford, GM, Motorola, and Hewlett Packard, were vertically integrated. A network of subcontractors fed into their manufacturing operations, but most of the assembly work was done in-house. Then in the late 1990’s, the twin phenomena of outsourcing and offshoring to reduce costs hollowed out the West’s electronics manufacturing infrastructure. The constant pressure to improve profit margins and quarterly financial results led OEMs to outsource many functions to lower-cost suppliers, except for wafer fabs, design, and intellectual property.
Today, the most critical factories in the world for high technology manufacturing all lie within 750 miles of China’s coastline. Almost every single OEM in the world now relies upon an Asia-centric supply chain.
But keep in mind: The issue is not just about China and the Asian tigers. The issue is competitiveness, and in many countries, competitive advantages are changing. In recent years, labor costs in Asia have risen as standards of living have improved. The cost to ship goods from Asia to North America or Europe is now five to 10 times higher than it was at the beginning of the COVID crisis, and it is projected to remain much higher for years. Rising demand is outstripping supplies.
All these factors are coming together to present an opportunity to re-establish a robust electronics supply chain in North America.
Silicon Valley was once the center of the electronics revolution that began in the 1970s. Quite often in those days, the entire supply chain from design to finished product was in California. If a part was needed, it was just a short drive away. These synergies created more wealth more rapidly than perhaps at any time in history.
The Taiwanese government took this lesson to heart, establishing the Silicon-Valley-inspired Hsinchu Science Park in 1980. Today, 400 companies across the electronics supply chain are once again located within a few miles of each other — but now in Taiwan. Major sub-component suppliers are all within a few hours’ drive, and suppliers work hand-in-glove with their top customers. Governments in China, South Korea, and elsewhere have replicated the model as well, putting materials, components, services, and suppliers all within easy reach. Government-backed capital and land deals played a supporting role.
With so much of the world’s most advanced electronics technologies now located in East Asia, the sector faces an existential crisis in the rest of the world. As we have seen over the last year, disruptions in the supply chain — whether caused by pandemics, political tensions, or other causes — could put entire national economies at risk.
The gaps in the North American electronics manufacturing supply chain are well known. The region lacks high-production manufacturing of the most advanced chips, IC packaging, and circuit boards. The circuit assembly and “box-build” industries are robust but not scaled to meet current global demand.
Leading-edge wafer fabs cost $10 billion or more. Investment in semiconductor fabrication is at a 20-year high, with Intel, TSMC, Samsung and others collectively committed to more than $50 billion in new U.S. wafer fabs. Because of the nature of their business, the electronics packaging suppliers, also known as Outsourced Semiconductor and Test (OSAT) suppliers, are closely involved with the wafer fabs from the outset. An OSAT factory can cost $300 million to $600 million.
This is where the CHIPS Act and the FABS Act stop. But as we’ve noted, it’s not where the industry stops. The allied printed circuit board (PCB) fabrication and assembly industries need support, too, if the United States wants to be a leader in building the most advanced electronic systems.
An advanced, high-production leading edge PCB factory can cost $400 million to $500 million. We have none in North America, and we would need several to become technology independent. Laser drills and direct imaging systems used in PCB manufacturing cost close to $1 million each, and hundreds are needed in high-production manufacturing.
Materials supply chains in the West have also contracted by 80% to 90% over the past 20 years. Some critical materials are no longer made in America. This must change as well.
The North American supply chain has reorganized itself to focus primarily on defense and aerospace, medical electronics, and other niche industries. But even in these critical sectors, the U.S. government admits there are serious technology gaps.
With rising automation, the Asian advantage of low-cost labor is reduced considerably. Also, the best and most critical technologies are still designed by Western companies. But the flip side is that American companies such as IBM, Apple and Microsoft may lead the world in intellectual property, but they farm out the manufacturing to subcontractors they don’t necessarily know very well.
To view it from another perspective, Apple, Microsoft, Qualcomm, and other IP-driven companies operate on gross margins of 35% to 65%. Chip makers like TSMC and Intel operate on 55% gross margins. An OSAT or PCB supplier typically operates at 10% to 15% margins, and an electronics manufacturing services (EMS) provider operates at 5% to 10% margins.
Thus, the competitiveness question comes down to long-range, public-private planning and collaboration; available working capital; tax and investment tradeoffs with overseas competition; and economies of scale. On all these measures, the United States has a great deal of work to do to be more competitive in the years ahead.
Strategically, electronics manufacturing is at the core of the modern economy. A strong and highly advanced technology sector is vital in a more crowded and dangerous world. From an economic perspective, electronics manufacturing lifts all boats.
But electronics supply chains have been and continue to be disrupted. The old models of offshoring no longer guarantee the availability of critical components.
Recovering from the mistakes of the last 20 years cannot happen without an integrated, collaborative plan. If America is to reclaim the mantle of global leadership on electronics manufacturing — and capture the benefits of doing so — it will need to focus on the whole ecosystem, not just one of its parts.
— Matt Holzmann is president of CGI Americas, based in Santa Ana, Calif.