A new era in computing may be upon us—one that doesn’t just challenge classical computing, but redefines how we think about processing power.
Meet the Thermodynamic Chip: A Radical Leap in Hardware
In a quiet lab outside of Boston, a groundbreaking chip has emerged. Developed by Extropic, a startup led by Guillaume Verdon, this thermodynamic chip is unlike anything currently on the market. Unlike traditional silicon chips or even quantum processors, this device leverages the chaotic nature of thermodynamic fluctuations to perform complex computations.
This innovation doesn’t fight randomness—it programs it. Verdon describes the chip as a generator of “programmable randomness,” designed to model uncertainty efficiently. This has wide-reaching implications, particularly in fields that rely on probabilistic simulations such as AI, financial modeling, and climate prediction.
Why Thermodynamics Over Quantum?
Verdon’s shift to thermodynamic computing wasn’t just a technical pivot—it was philosophical. After working on quantum computing at Google and contributing to TensorFlow Quantum, Verdon became disillusioned with the slow progress and limitations of quantum systems. He saw thermodynamics not as a hindrance, but as untapped potential.
Rather than cooling systems to near absolute zero to remove noise—as is required in quantum computing—thermodynamic hardware embraces heat and entropy. This not only simplifies the hardware but could also drastically cut energy costs. The approach marks a bold departure from conventional computing paradigms, offering a new frontier that could eventually run entire AI workloads natively.
Accelerating Innovation with e/acc
Verdon is also the outspoken voice behind the online persona Based Beff Jezos, a leading figure in the ideology of effective accelerationism (e/acc). This movement rejects the caution urged by AI safety advocates and champions rapid, unconstrained technological advancement. It’s a philosophy that has gained traction among tech leaders and investors seeking to push innovation forward without regulatory drag.
In the e/acc view, technological evolution is inevitable and should be embraced with open arms. It frames the universe as biased toward greater complexity and intelligence, and humans—as well as AI—are simply expressions of that natural tendency to extract and utilize energy more efficiently.
From Concept to Market—Fast
Extropic’s first chip was conceived less than a year ago, yet the company is already planning to deliver operational hardware by the end of 2025. This pace is extraordinary, especially given the complexity of building a new kind of processor from the ground up.
Verdon and his co-founder Trevor McCourt—both former quantum computing researchers—are leading a team of around 20 engineers. Backed by $14.1 million in seed funding, their goal is to bring thermodynamic computing into real-world applications like high-frequency trading and medical research, where modeling uncertainty is critical.
Quantum Disillusionment Fuels Thermo Revolution
The pivot away from quantum computing isn’t unique to Verdon. Other startups, like Normal Computing, are also exploring thermodynamic approaches after becoming skeptical about the commercial viability of noisy quantum processors. The common theme: quantum promises are maturing too slowly, and the challenges of error correction and extreme cooling remain formidable hurdles.
Meanwhile, the explosion of generative AI has intensified the demand for more efficient computing architectures. As AI systems become more complex, traditional chips are struggling to keep pace—fueling interest in alternative paradigms like thermodynamic computing. This transition aligns well with the broader AI revolution, where even developers are increasingly relying on new AI tools to reshape how software is built.
Bringing Chaos to Order: A New Computing Philosophy
Extropic’s chips are not just about hardware—they represent a shift in how we conceptualize computing itself. Instead of suppressing the randomness inherent in physical systems, this new model harnesses it. Under a microscope, the chip reveals a delicate lattice of components, each designed to allow probabilities—not certainties—to drive computations.
As Verdon puts it, this is about embracing the “will of the universe”—a thermodynamic path toward more intelligent, energy-efficient civilizations. It’s a vision where computing systems evolve like life itself: by exploiting energy gradients to achieve higher-order functionality.
The Road Ahead
With early adopters in finance and biotech eager to test Extropic’s thermodynamic accelerators, 2025 could mark a turning point. If successful, these chips may prove not only faster but more sustainable than their classical or quantum counterparts.
From the ashes of quantum skepticism and the rise of AI’s computational hunger, a new paradigm is forming—one rooted in the laws of physics and driven by a philosophy of acceleration. Whether this bold vision will reshape the future of computing remains to be seen, but one thing is certain: the thermodynamic revolution has begun.
