Our Take
Solid-state cooling is real and in the field, but the efficiency problem is unsolved; the honest play is niche deployment, not AC replacement.
Why it matters
The IEA projects AC units will triple by 2050, and current systems use refrigerants with global-warming potential 2,000x that of CO2. Even a small market shift to lower-GWP alternatives would cut emissions meaningfully.
Do this week
Climate tech investors: monitor Mimic's Vancouver pilot and Magnotherm's supermarket test results over the next 12 months before committing capital to solid-state cooling bets.
Four competing solid-state cooling systems entered pilot testing
Brooklyn-based Mimic Systems is running a thermoelectric room-scale climate control pilot in a Vancouver apartment. The German company Magnotherm is testing a magnetocaloric system in supermarket chains. A Hong Kong team demonstrated an elastocaloric device reaching below 0°C. The UK's Barocal is piloting a barocaloric approach. All four replace traditional compressor-and-refrigerant AC with passive conductive materials (gadolinium, bismuth telluride) or active material-state changes (magnetization, expansion, pressure shifts) to move heat.
The pressure to move is real: the International Energy Agency projects AC units will triple by 2050 (company-reported). A 2019 Lancet study estimated AC prevented nearly 200,000 premature deaths that year alone. But current AC uses refrigerant R410A, which carries global-warming potential 2,000+ times that of CO2.
Efficiency remains the hard problem
Conventional HVAC systems achieve a coefficient of performance (COP) of around 3—meaning they move three units of heat per unit of electrical energy input. Thermoelectric systems, the furthest along technically, perform much worse at high temperature differentials, limiting them to niche applications like cooling car-seat backs, says Jeff Snyder at Northwestern University.
Mimic claims its room-scale unit should match a typical AC's annual energy draw, but no independent benchmarks yet exist to confirm this across a full cooling season. Elastocaloric and barocaloric systems show promise but remain two to three years from room-scale prototype maturity (per Lindsay Rasmussen, Rocky Mountain Institute). The crux: without long-term energy consumption data from deployed units, the efficiency question cannot be answered.
Some factors favor solid-state anyway. These systems have fewer moving parts, which should improve durability compared to compressor-based units prone to mechanical wear. And they eliminate the high-GWP refrigerant problem. But durability gains and refrigerant benefits do not offset efficiency losses if they cost more electricity to operate.
Replacement is unlikely; niche penetration is the realistic target
Solid-state cooling will not supplant traditional AC at scale. The efficiency gap is too wide, and the installed base of conventional systems is enormous. But Rasmussen notes that if solid-state technologies capture even 5% market share, that represents "a really large potential impact" as India alone is set to install tens of millions of new AC units over the next decade.
The proof will come from the field. Mimic and Magnotherm data over the next 12–24 months will answer whether thermoelectric and magnetocaloric systems can close the efficiency gap in real operating conditions. Elastocaloric and barocaloric systems require longer R&D runway but may offer better performance per watt if the physics scales. Until then, the honest story is not replacement—it is whether solid-state can carve a profitable wedge in high-value use cases (data centers, EV thermal management, cold storage) before attempting the mainstream residential and commercial markets.