Wednesday, September 16, 2026
Tech Beat
Sep 16, 2026, 12:47 PMArtificial Intelligence

Syensqo Uses Microsoft AI to Build Next Generation Data Center Materials

Syensqo uses Microsoft AI to screen millions of molecules as it develops chips, cooling and high-voltage materials for more efficient AI infrastructure.

Listen to this briefingAudio briefing

Summary

Syensqo outlined on September 16, 2026, how AI is pushing semiconductors and data centers toward limits in heat, electrical efficiency, purity and reliability. Mike Finelli, chief technology and innovation officer and chief North America officer, said the company is developing high-voltage data center materials, plasma-resistant seals for wafer tools and dielectric fluids for direct immersion cooling. Higher voltage can reduce electrical losses but increases power density and temperature. Syensqo is adapting electric vehicle insulation designed to carry 100 kilowatts during rapid acceleration to 60 miles per hour, alongside a lithium-ion cathode binder designed for 10-year battery life, for data center power distribution, renewable energy storage and backup power.

Finelli, who joined Syensqo 33 years ago, said 20% of annual revenue comes from products and applications launched within five years. Syensqo classifies 88% of its portfolio as sustainable through its Sustainable Portfolio Management tool, which evaluates production impact and customer benefits before research begins. It is also developing heat-transfer fluids with lower global-warming impact than existing options.

For two years, Syensqo has worked with Microsoft Discovery and AI agents that digitally synthesize millions or tens of millions of molecular combinations. Physics-based simulations predict physical and chemical performance, toxicity and sustainability, then rank roughly 100 candidates for laboratory synthesis. Syensqo says this accelerates discovery without replacing scientists, potentially creating a feedback loop in which AI develops materials that improve AI infrastructure, although its speed remains uncertain.

Positives

  • 20% of Syensqo’s annual revenue comes from products and applications introduced during the past five years.
  • 88% of Syensqo’s portfolio qualifies as sustainable under its Sustainable Portfolio Management assessment.
  • Millions of potential molecules can be reduced to roughly 100 priority candidates for laboratory synthesis.
  • High-voltage architectures could increase computing power while reducing data center electrical losses and environmental impact.
  • Electric vehicle insulation, battery binders and cooling expertise are being transferred into AI data center infrastructure.

Risks & concerns

  • Higher-voltage data centers increase power density and temperatures even as they improve electrical efficiency.
  • Existing heat-transfer fluids can carry high global-warming impacts, requiring lower-impact replacements.
  • Semiconductor seals face rising temperatures, aggressive plasmas, reactive chemicals, outgassing and purity requirements.
  • AI simulations provide predicted performance, leaving roughly 100 shortlisted molecules requiring physical laboratory synthesis.
  • The speed of the proposed feedback loop between better AI and faster materials discovery remains uncertain.
Primary sourceArtificial intelligence – MIT Technology Reviewhttps://www.technologyreview.com/2026/09/16/1144014/building-the-materials-foundation-for-ai/
Read full article
Editorial note: Tech Beat summarizes and analyzes third-party reporting. The source link is the authoritative article. This page does not reproduce the full source text.

More From The Wire

Artificial IntelligenceSep 16

OpenAI and AARP Bring Free ChatGPT Training to 1,000 Older Adults

Artificial IntelligenceSep 16

Amazon Alexa+ Launches in India With Hindi and ₹2,000 Non-Prime Price

Artificial IntelligenceSep 16

University of Manchester Uses NVIDIA Earth-2 for UK Pollution Forecasts