Phonon Focusing at Room Temperature: What It Means for AI Chips

💡 For the first time, UCLA engineers demonstrated in Nature Physics (July 2026) that phonon focusing - where heat travels in focused, directed rays - can occur at room temperature. Previously confined to near absolute zero, this opens a path to routing heat in chips as precisely as engineers already route electricity or light.

What is a phonon, and why does heat scatter instead of focusing?

Heat in a solid is carried by phonons: quantized vibrations that ripple through a crystal's atomic lattice. Think of them as the microscopic packets of thermal energy, analogous to photons for light. In most materials, phonons collide constantly and scatter in all directions, making heat diffuse outward like gas in a room - impossible to direct or concentrate.

In certain structured crystals, phonons can instead stay coherent and travel in focused, ray-like paths. The shape of those rays is set by the crystal geometry. For decades, this phenomenon called phonon focusing was observed only at cryogenic temperatures near 0 Kelvin, because at room temperature, thermal noise disrupts phonon coherence before any focusing can form.

What did the UCLA team actually discover?

Professor Yongjie Hu and his group at the UCLA Samueli School of Engineering chose boron arsenide, a crystalline semiconductor with unusually weak phonon scattering. Using nanoscale temperature-mapping techniques, they measured heat flow at 300 K (about 27°C, room temperature) and observed ray-like focusing for the first time outside a cryostat.

Depending on which crystal plane they studied, the heat formed sixfold, eightfold, or fourfold symmetry patterns. The coherent beam persisted over at least 1 micrometer, with projections extending to tens of micrometers. The results were published in Nature Physics on July 23, 2026.

What does this mean for AI chips and quantum hardware?

Heat is the hard ceiling on chip density. As AI accelerators pack more transistors per square millimeter, the thermal load per unit area grows faster than cooling solutions can keep up. Today's approaches - liquid cooling, vapor chambers, diamond heat spreaders - deal with heat after it has spread and randomized. They are good solutions to a hard problem, but they do not change the physics of how heat moves through the chip itself.

Phonon focusing would change that physics. If heat can be designed to flow along specific routes, away from computation cores and toward heat sinks, chips could run hotter in aggregate while keeping sensitive areas cool. For quantum computers specifically, stray heat causes decoherence, which destroys the quantum state. Redirecting heat at the nanoscale addresses that problem directly.

This sits alongside other challenges in AI infrastructure energy management, like advances in fuel cell catalysts for data centers, though the underlying physics are entirely different.

Is boron arsenide ready for real chips?

Not yet, and possibly not for years. Several real obstacles stand in the way:

• Material availability: Boron arsenide is not a mainstream semiconductor. Growing large, defect-free crystals is difficult, and no wafer-scale manufacturing supply chain exists.
• Device integration: Demonstrating phonon focusing in an isolated crystal is very different from embedding it in a multilayer chip with electrical interconnects, dielectric layers, and metal contacts.
• Scale: 1-10 micrometers of coherent heat routing is promising but modest compared to the millimeter-to-centimeter distances that matter for full-chip thermal management.
• Competition: Diamond heat spreaders, liquid metal interfaces, and 3D vapor chambers are already deployed in high-performance chips. Phonon routing needs a clear manufacturing advantage, not just a lab effect.

Professor Hu describes this as "establishing a foundation for quantum thermal engineering" - an honest framing. The key advance is that cryogenic cooling is no longer a prerequisite for studying this effect.

What comes next, and when could this reach real chips?

The next meaningful milestones would be: observing phonon focusing in silicon or gallium nitride (mainstream chip materials), extending coherence length, and showing that focusing direction can be designed into the crystal during growth rather than discovered after. Any of those steps would shift this from fundamental physics to materials engineering for real devices.

The broader pattern worth noting: 2026 is seeing several effects once limited to cryogenic conditions move to ambient temperature. Each such crossing expands what engineers can actually build. The year has also brought other advances in observing previously invisible physical phenomena at the frontier of measurement science.

FAQ

Sources: UCLA Newsroom - Engineers observe quantum heat waves at room temperature (2026); ScienceDaily - UCLA scientists discover how to guide heat like light at room temperature (August 2026)

About the author

Dao Huy (Lucas) is a professional translator working across English, Vietnamese, Chinese, and French. He follows scientific and technical frontiers not as an engineer or investor, but as someone fascinated by how new ideas need to be explained clearly before they can matter - and how much precision that explanation requires.

He offers technical, scientific, and IP translation services (English to Vietnamese, Chinese, and French) and software localization. If your research, product, or company needs precise multilingual communication, he is happy to help - visit daohuy.com for a quote.