Snowcap raises $23M for superconducting AI chips promising 25x efficiency gains

Snowcap Compute has raised $23 million to develop superconducting AI chips that could dramatically outperform current systems while consuming far less electricity. The startup claims its technology will be 25 times more efficient than today’s best chips in performance per watt, even after accounting for the energy needed to keep the superconducting materials at extremely cold temperatures.

The big picture: As AI systems demand increasingly massive amounts of power—with Nvidia‘s upcoming “Rubin Ultra” server expected to consume 600 kilowatts—the industry is desperately seeking alternatives to conventional silicon chips that are hitting physical limits.

How it works: Superconductors are materials that allow electrical current to flow without resistance, but they only function at cryogenic temperatures requiring specialized cooling systems.

• While scientists have theorized about superconducting computer chips since the 1990s, the cooling requirements previously made them impractical due to high energy consumption.
• The explosive growth in AI computing demand has changed the calculus, making dedicated cooling infrastructure worthwhile if performance gains are substantial enough.
• The chips can be manufactured in standard factories but require exotic materials like niobium titanium nitride, which depends on ingredients from Brazil and Canada.

Key details: Snowcap’s founding team brings deep expertise from across the semiconductor industry and defense sectors.

• The team includes Anna Herr and Quentin Herr, scientists with extensive superconducting chip experience at research firm Imed and defense contractor Northrop Grumman.
• Former chip executives from Nvidia and Google’s Alphabet round out the leadership team.
• CEO Michael Lafferty previously oversaw futuristic chip development at Cadence Design Systems, a software company that helps design computer chips.

Timeline and backing: The company plans to deliver its first basic chip by the end of 2026, with full systems arriving later.

• Pat Gelsinger, Intel’s former CEO, led the investment for venture firm Playground Global and is joining Snowcap’s board.
• Cambium Capital and Vsquared Ventures also participated in the funding round.

What they’re saying: Industry leaders emphasize the urgent need for more power-efficient computing solutions.

• “Power (efficiency) is nice, but performance sells,” Lafferty said. “So we’re pushing the performance level way up and pulling the power down at the same time.”
• “A lot of data centers today are just being limited by power availability,” Gelsinger noted, highlighting the need for a “sharp break” from the current trajectory of ever-increasing electricity consumption.

Why this matters: Current AI infrastructure is straining electrical grids, with single high-end servers consuming as much power in one hour as an average U.S. household uses in two-thirds of a month, making revolutionary efficiency improvements critical for the industry’s sustainable growth.