Toward Large-Scale Photonics-Empowered AI Systems: From Physical Design Automation to System-Algorithm Co-Exploration

2025-12-31

AI Summary

This work addresses key challenges in large-scale photonic AI systems, including limited support for dynamic tensor operations such as attention mechanisms, high overhead from optoelectronic conversion and data movement, and degraded robustness due to non-idealities at high integration densities. To overcome these limitations, the authors propose the first end-to-end photonic AI design framework that enables dynamic tensor computation, co-optimizes photonic and electronic components, and incorporates manufacturing constraints to facilitate joint algorithm–system–physical layer exploration. Leveraging SimPhony-based modeling, ADEPT circuit topology optimizations, and Apollo/LiDAR-driven physical design automation, the framework holistically accounts for routing, thermal effects, and crosstalk, yielding a manufacturable, energy-efficient, and high-density photonic AI system. The feasibility of the approach is validated under realistic hardware constraints.

Technology Category

• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV
• Machine Learning: Hardware-aware ML

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

In this work, we identify three considerations that are essential for realizing practical photonic AI systems at scale: (1) dynamic tensor operation support for modern models rather than only weight-static kernels, especially for attention/Transformer-style workloads; (2) systematic management of conversion, control, and data-movement overheads, where multiplexing and dataflow must amortize electronic costs instead of letting ADC/DAC and I/O dominate; and (3) robustness under hardware non-idealities that become more severe as integration density grows. To study these coupled tradeoffs quantitatively, and to ensure they remain meaningful under real implementation constraints, we build a cross-layer toolchain that supports photonic AI design from early exploration to physical realization. SimPhony provides implementation-aware modeling and rapid cross-layer evaluation, translating physical costs into system-level metrics so architectural decisions are grounded in realistic assumptions. ADEPT and ADEPT-Z enable end-to-end circuit and topology exploration, connecting system objectives to feasible photonic fabrics under practical device and circuit constraints. Finally, Apollo and LiDAR provide scalable photonic physical design automation, turning candidate circuits into manufacturable layouts while accounting for routing, thermal, and crosstalk constraints.

Problem

• photonic AI systems
• dynamic tensor operations
• hardware non-idealities
• data-movement overheads
• large-scale integration

Innovation

• photonic AI
• system-algorithm co-exploration
• physical design automation
• dynamic tensor operations
• hardware non-idealities