'Thermodynamic computer' can mimic AI neural networks — using orders of magnitude less energy to generate images
livescience.com
February 23, 2026
8 min read
🔥🔥🔥🔥🔥
54/100
Summary
A thermodynamic computer can generate images from noise while consuming significantly less energy than traditional generative AI models. This technology mimics the functionality of AI neural networks.
Key Takeaways
- Researchers developed a "thermodynamic computer" that generates images from thermal noise, significantly reducing energy consumption compared to traditional AI models.
- This new computing method leverages random fluctuations in energy, allowing it to perform tasks with orders of magnitude less energy than current generative AI systems.
- The thermodynamic computer operates using probabilities rather than definite binary values, enhancing efficiency for optimization problems.
- The findings were published in the journal Physical Review Letters on January 20.