Key Summary

• AI materials discovery is gaining momentum as chipmakers look to ease semiconductor supply constraints.
• AI memory demand remains tight, with SK Hynix’s HBM position signaling continued pressure.
• India is building a broader semiconductor ecosystem, not just subsidizing factories.

by Carrie Sessine, Chief Communications Officer, Global Electronics Association

Three stories, one thread: where the electronics industry’s next constraints are being decided.

I appreciate the chance we have to share what we’re reading and watching and how these stories shape market developments and ripple through the electronics industry. I’m back again with three that stuck with me this month.

Here’s the red-thread theme: the same story is playing out three times right now. Capital and policy are racing to lock down the physical building blocks of electronics: the materials, the memory, and the manufacturing. Read together, they tell you where our industry’s next constraints, and next advantages, are being decided.

AI takes on the hardest problem in our industry: materials

The story: CuspAI, which calls itself a “structured search engine for physical materials discovery,” raised a $450M Series B co-led by Kleiner Perkins and NEA reaching a $2.6B valuation and more than half a billion dollars raised since it launched in 2024.

Why it matters: Some of the toughest bottlenecks in semiconductor manufacturing are materials problems such as compound substrates. CuspAI is betting AI can compress decades of materials R&D into months.

The tell: AMD Ventures joined the round, and Samsung, Nvidia and Lam Research signed on to CuspAI’s new “AI Materials Foundry.” When chipmakers back your search engine, they’re pointing straight at the constraint.

What I’m watching: CuspAI’s founder says semiconductors will take up 80% of its research this year, including work to replace supply-constrained metals like ruthenium and iridium. That makes this a supply-chain story as much as an AI one.