Technology Category
Computer Vision: Multi-modal VisionPhilosophy and Ethics of AI: Philosophical Foundations of AIPlanning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
High-speed multiwavelength photonic temporal integration using silicon photonics
🤖 AI Summary
A fundamental bottleneck exists in photonic AI between the difficulty of large-vector mapping, slow thermal-optic response, and the demand for ultrafast computation. Method: This paper proposes a time-unrolled, multi-wavelength photonic temporal integration architecture. It introduces an on-chip photonic heating-integration layer (PHIL), which counterintuitively exploits thermal dissipation delay to achieve all-optical temporal integration of 50-GHz-speed optical signals. The architecture unifies linear and nonlinear all-optical operations on a single silicon photonic integrated platform, eliminating electro-optic conversion overhead. Contribution/Results: By synergistically leveraging thermo-optic tuning, multi-wavelength multiplexing, and photonic temporal signal unfolding, the approach experimentally demonstrates high-fidelity, high-bandwidth (50 GHz) all-optical integration and end-to-end signal processing. This work establishes a new paradigm for scalable, high-throughput, low-power photonic AI acceleration.
Technology Category
Application Category
📝 Abstract
Optical systems have been pivotal for energy-efficient computing, performing high-speed, parallel operations in low-loss carriers. While these predominantly analog optical accelerators bypass digitization to perform parallel floating-point computations, scaling optical hardware to map large-vector sizes for AI tasks remains challenging. Here, we overcome this limitation by unfolding scalar operations in time and introducing a photonic-heater-in-lightpath (PHIL) unit for all-optical temporal integration. Counterintuitively, we exploit a slow heat dissipation process to integrate optical signals modulated at 50 GHz bridging the speed gap between the widely applied thermo-optic effects and ultrafast photonics. This architecture supports optical end-to-end signal processing, eliminates inefficient electro-optical conversions, and enables both linear and nonlinear operations within a unified framework. Our results demonstrate a scalable path towards high-speed photonic computing through thermally driven integration.
Problem
Research questions and friction points this paper is trying to address.
scaling optical hardware for large-vector AI tasks
bridging speed gap between thermo-optic effects and ultrafast photonics
enabling linear and nonlinear optical operations without electro-optical conversions
Innovation
Methods, ideas, or system contributions that make the work stand out.
PHIL unit enables all-optical temporal integration
Slow heat dissipation integrates 50 GHz signals
Unified framework for linear and nonlinear operations
🔎 Similar Papers
No similar papers found.
