$UMC is my long-term semiconductor play

Not financial advice. Just sharing how I’m looking at United Microelectronics Corporation and why I think it’s more interesting than most people realize.

Everyone talks about AI chips and cutting-edge nodes, but most chips in the world don’t need that. UMC focuses on mature and specialty nodes used in the automobile industry, industrial equipment, power management chips, displays, IoT devices, and networking gear. These markets care more about stability, reliability, and cost than raw performance.

Once a customer designs a chip on UMC’s process, they usually stick around for years which creates consistent and predictable demand. Over 80% of global chip volume still comes from mature nodes, and that market is not disappearing anytime soon. With expansions in Singapore, UMC wafer volume capacity is set to be 8 million+ in 2026. This is about half of that of TSMC, but more than other leading global mature node foundries.

UMC is not frozen in the past. The company offers 14nm FinFET technology, which gives customers approximately 50–60% better power efficiency versus 28nm, along with improved performance at a reasonable cost. This is ideal for companies that want better chips but don’t want to pay premium prices for leading-edge production. UMC does not need to beat TSMC. It just needs to be dependable and it is.

Intel is collaborating with UMC on a 12nm process expected to be manufactured in the United States later this decade. This matters because customers increasingly want less geographic concentration risk. UMC gains access to U.S. manufacturing capacity without building massive new fabs, and Intel benefits from UMC’s experience running a high-mix foundry business. If even 5–10% of UMC’s revenue eventually comes from U.S. manufacturing, that alone could justify a higher valuation.

It’s not TSMC, but the market prices UMC as if growth is basically zero. A 20–30% re-rating over the next few years does not seem unreasonable if utilization improves, the Intel partnership delivers, and mature-node demand remains steady.

UMC is not a moonshot, but it’s more than a boring chip foundry that keeps the world running. This company is trading at a depressed valuation, with upside optionality tied to U.S. manufacturing and supply-chain diversification.

I also believe that this is going to be their year. Volume has been moving like there’s something about to be unearthed.

It may not be exciting, but boring businesses with low expectations can still make money if they keep executing. UMC has been doing exactly that for a long time.