Analog optical computer for AI inference and combinatorial optimization

• 成果类型: Article
• 署名作者: Kalinin, Kirill P.; Gladrow, Jannes; Chu, Jiaqi; Clegg, James H.; Cletheroe, Daniel; Kelly, Douglas J.; Rahmani, Babak; Brennan, Grace; Canakci, Burcu; Falck, Fabian; Hansen, Michael; Kleewein, Jim; Kremer, Heiner; O'Shea, Greg; Pickup, Lucinda; Rajmohan, Saravan; Rowstron, Ant; Ruhle, Victor; Braine, Lee; Khedekar, Shrirang; Berloff, Natalia G.; Gkantsidis, Christos; Parmigiani, Francesca; Ballani, Hitesh
• 署名单位: Microsoft; Microsoft United Kingdom; Microsoft; University of Cambridge
• 刊物名称: NATURE
• ISSN/ISSBN: 0028-0836; 1476-4687
• DOI: 10.1038/s41586-025-09430-z
• 发表日期: 2025-09-01
• 页码: 354-+
• 关键词: artificial-intelligence

摘要

Artificial intelligence (AI) and combinatorial optimization drive applications across science and industry, but their increasing energy demands challenge the sustainability of digital computing. Most unconventional computing systems(1-7) target either AI or optimization workloads and rely on frequent, energy-intensive digital conversions, limiting efficiency. These systems also face application-hardware mismatches, whether handling memory-bottlenecked neural models, mapping real-world optimization problems or contending with inherent analog noise. Here we introduce an analog optical computer (AOC) that combines analog electronics and three-dimensional optics to accelerate AI inference and combinatorial optimization in a single platform. This dual-domain capability is enabled by a rapid fixed-point search, which avoids digital conversions and enhances noise robustness. With this fixed-point abstraction, the AOC implements emerging compute-bound neural models with recursive reasoning potential and realizes an advanced gradient-descent approach for expressive optimization. We demonstrate the benefits of co-designing the hardware and abstraction, echoing the co-evolution of digital accelerators and deep learning models, through four case studies: image classification, nonlinear regression, medical image reconstruction and financial transaction settlement. Built with scalable, consumer-grade technologies, the AOC paves a promising path for faster and sustainable computing. Its native support for iterative, compute-intensive models offers a scalable analog platform for fostering future innovation in AI and optimization.