AI is pushing computing into new territory, but the next breakthrough may not come from an algorithm. It may come from the materials surrounding the chips, power systems, cooling equipment, and data centers that make modern AI possible.
That foundation is under pressure. Semiconductors and data centers are approaching physical limits around performance, thermal management, electrical efficiency, and reliability, while the physical infrastructure behind AI growth cannot move at the same pace as demand. With ChatGPT recently reaching 1 billion monthly app users, the need for stronger infrastructure is becoming impossible to ignore.
AI Infrastructure Is Running Into Physical Limits
“AI is now, from a material standpoint, really pushing semiconductors and the data centers to their physical limits,” said Mike Finelli, chief technology and innovation officer and chief North America officer at Syensqo.
Those limits reach across every layer of the system. Higher performance creates tougher thermal-management demands. Greater electrical needs place new pressure on efficiency and reliability. Semiconductor manufacturing also requires materials that can survive demanding chemical and plasma environments without losing performance over long periods.
As these requirements rise, materials move toward what Finelli calls the “top of the pyramid.” Reaching that level requires a combination of high temperature performance, purity, electrical performance, chemical resistance, plasma resistance, and long-term stability. A material that succeeds in one area but fails in another cannot support the next generation of AI infrastructure.
That shift gives advanced materials a defining role. They are no longer just components inside larger systems; they increasingly determine what AI infrastructure can do, how much energy it can use, and how long it can operate.
New Materials Target Power, Cooling, and Manufacturing
Syensqo is developing materials for high-voltage data center architectures, advanced sealing materials for semiconductor manufacturing, and thermal-management solutions that include fluids for direct immersion cooling. Each area addresses a pressure point created by AI’s demand for more computing capacity.
High-voltage architectures can help data centers handle higher electrical demands, while advanced sealing materials support the challenging conditions involved in semiconductor manufacturing. Direct immersion cooling takes aim at heat, using fluids to manage thermal loads around computing equipment.
Materials developed for electric vehicles can also help address higher voltage and energy-density demands in data centers. The connection matters because both systems face pressure to move more energy through compact infrastructure while maintaining performance and reliability.
Customers are also asking for more than technical performance. More customers expect materials to meet demanding specifications while reducing environmental impact, creating a challenge that reaches beyond speed, power, or durability.
“Our goal is to remove the trade-off between performance and sustainability,” Finelli said.
AI Is Becoming a Materials Discovery Engine
The same technology driving demand for advanced materials is also helping scientists find them. Materials scientists are using AI to search the universe of possible molecules and accelerate the development of new solutions.
Syensqo is using AI agents to digitally synthesize millions of potential molecular combinations, predict their performance and sustainability, and narrow the field for laboratory testing. Instead of testing every possibility one by one, researchers can use AI to identify candidates that meet specific technical and environmental requirements before those candidates reach the lab.
This process connects digital discovery with physical manufacturing. AI explores molecular combinations, predicts how they may perform, and helps focus laboratory testing on the most promising options. The result is a tighter loop between computation, experimentation, and product development.
That loop also creates a powerful cycle for AI itself. AI helps develop materials that improve AI infrastructure, and that stronger infrastructure enables better AI to accelerate materials discovery. “You end up in this accelerated materials, innovative cycle of materials innovation,” Finelli said.
The stakes extend beyond one company or one type of equipment. Semiconductors, data centers, cooling systems, high-voltage architectures, and manufacturing processes all depend on materials that can meet rising demands at the same time.
AI’s future will therefore be shaped by two connected searches: the search for more capable algorithms and the search for materials that let those algorithms run at scale. As computing approaches its physical boundaries, the materials foundation will decide how far the next wave of AI can go.
Based on
