CrossLight: A Cross-Layer Optimized Silicon Photonic Neural Network Accelerator

Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and performance compared to CPUs and GPUs. In this paper, we propose a novel cross-layer optimized neural network accelerator called CrossLight that leverages silicon photon...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2021 58th ACM/IEEE Design Automation Conference (DAC) s. 1069 - 1074
Hlavní autoři: Sunny, Febin, Mirza, Asif, Nikdast, Mahdi, Pasricha, Sudeep
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 05.12.2021
Témata:
Crosstalk
deep learning
Neural networks
Performance evaluation
Power lasers
Silicon photonics
Thermal engineering
Throughput
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and performance compared to CPUs and GPUs. In this paper, we propose a novel cross-layer optimized neural network accelerator called CrossLight that leverages silicon photonics. CrossLight includes device-level engineering for resilience to process variations and thermal crosstalk, circuit-level tuning enhancements for inference latency reduction, and architecture-level optimizations to enable better resolution, energy-efficiency, and throughput. On average, CrossLight offers 9.5x lower energy-per-bit and 15.9x higher performance-per-watt than state-of-the-art photonic deep learning accelerators.
DOI:10.1109/DAC18074.2021.9586161