Samsung Electronics says it can use artificial intelligence to dramatically shorten the time it takes to develop new materials for making semiconductors — from five years down to two. The plan, shared at an industry forum in South Korea by Choi Young-sang of the Samsung Advanced Institute of Technology, the company's research arm, was reported by Yonhap. It's a bet that AI can do the heavy lifting in the slowest, most tedious part of materials science: trial and error.
In chip manufacturing, the materials used to build each layer of a chip — from insulating films to conductive metals — have to meet extremely precise chemical and physical standards. Finding the right combination has traditionally meant running thousands of physical experiments in a lab, each one taking weeks or months. If a material fails, you learn from it and try again. That process is why bringing a new material from concept to production can take half a decade.
Samsung says AI can change that equation. By virtually screening more than 100 million possible material combinations for a seven-layer chip structure, the AI can predict which ones are likely to work before any physical testing happens. It can also learn from failed experiments, feeding that information back into the model to avoid repeating dead ends. According to Samsung, this approach could cut the number of real-world lab runs needed by more than 80%.
If that holds up, the payoff is significant. Samsung says the method could surface more than five times as many promising material candidates while lowering development costs. And the same technique could later be applied to other hard-to-test areas, such as battery materials, which face similar trial-and-error bottlenecks.
Why R&D speed matters for a chip giant
For a company like Samsung, which spends billions of dollars a year on semiconductor research, the speed of development is almost as important as the outcome. Faster materials development means new chip designs can move into production sooner, and production processes can be refined more quickly. It also reduces the risk that a promising material will hit an unexpected problem late in the qualification process, which can cause costly delays and production hiccups.
This is especially relevant in the current AI-driven chip boom. Demand for advanced chips — particularly those used in AI data centers — has surged, and chipmakers are racing to improve performance and efficiency. Materials are often the limiting factor in how small, fast, and power-efficient a chip can be. Being able to test more candidates in less time gives Samsung a potential edge over competitors.
The move also fits a broader trend in the industry. AI chip rallies have lifted Samsung and SK Hynix recently, as investors bet on the long-term demand for advanced semiconductors. Samsung has also been expanding its chip substrate capacity, with Nvidia and Tesla reportedly backing a $2.8 billion expansion. Faster materials development could support those efforts by ensuring a steady supply of qualified materials for new production lines.
What it means for investors
For everyday investors, Samsung's two-year target is really a bet on R&D throughput. If AI can reliably eliminate most dead-end experiments, Samsung can run many more “shots on goal” per dollar of research spending and move winners into manufacturing sooner. That changes the risk profile of its semiconductor pipeline: progress depends less on a single big materials bet working out, and more on a wider funnel of viable options. Over time, that can mean steadier production ramps and fewer surprise cost spikes when a material hits problems late in qualification.
It's important to note that this is an ambition, not a guarantee. AI-driven materials discovery has shown promise in academic and industrial settings, but scaling it to the complexity of real chip manufacturing is a different challenge. The 80% reduction in lab runs and the two-year timeline are targets, not results. Investors should watch for evidence that Samsung is actually hitting those milestones in its production lines.
Still, the announcement underscores how seriously Samsung is taking AI as a tool for its own operations, not just as a product feature. The company has been integrating AI into its chip manufacturing and packaging, and this materials initiative is another piece of that strategy.
For the broader market, the story is a reminder that AI's impact goes beyond chatbots and data centers. It's also becoming a productivity tool for the companies that make the hardware AI runs on. If Samsung can compress its R&D cycles, it could set a precedent that other chipmakers and materials firms follow, potentially accelerating innovation across the entire semiconductor supply chain.
Investors in Samsung — or in the wider semiconductor sector — will be watching to see whether these AI-driven efficiencies translate into faster product launches, better margins, and fewer supply disruptions. For now, the plan is a promising sign, but the proof will be in the production line.
