skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: PCM Enabled Low-Power Photonic Accelerator for Inference and Training on Edge Devices
The convergence of edge computing and artificial intelligence requires that inference is performed on-device to provide rapid response with low latency and high accuracy without transferring large amounts of data to the cloud. However, power and size limitations make it challenging for electrical accelerators to support both inference and training for large neural network models. To this end, we propose Trident, a low-power photonic accelerator that combines the benefits of phase change material (PCM) and photonics to implement both inference and training in one unified architecture. Emerging silicon photonics has the potential to exploit the parallelism of neural network models, reduce power consumption and provide high bandwidth density via wavelength division multiplexing, making photonics an ideal candidate for on-device training and inference. As PCM is reconfigurable and non-volatile, we utilize it for two distinct purposes: (i) to maintain resonant wavelength without expensive electrical or thermal heaters, and (ii) to implement non-linear activation function, which eliminates the need to move data between memory and compute units. This multi-purpose use of PCM is shown to lead to significant reduction in energy consumption and execution time. Compared to photonic accelerators DEAP-CNN, CrossLight, and PIXEL, Trident improves energy efficiency by up to 43% and latency by up to 150% on average. Compared to electronic edge AI accelerators Google Coral which utilizes the Google Edge TPU and Bearkey TB96-AI, Trident improves energy efficiency by 11% and 93% respectively. While NVIDIA AGX Xavier is more energy efficient, the reduced data movement and GST activation of Trident reduce latency by 107% on average compared to the NVIDIA accelerator. When compared to the Google Coral and the Bearkey TB96-AI, Trident reduces latency by 1413% and 595% on average.  more » « less
Award ID(s):
2324644 2324645 1901192 1936794 1901165
PAR ID:
10547994
Author(s) / Creator(s):
; ; ;
Publisher / Repository:
IEEE
Date Published:
ISBN:
979-8-3503-6460-6
Page Range / eLocation ID:
600 to 607
Format(s):
Medium: X
Location:
San Francisco, CA, USA
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; (, IEEE Transactions on Sustainable Computing)
    The growing computational demands of deep learning have driven interest in analog neural networks using resistive memory and silicon photonics. However, these technologies face inherent limitations in computing parallelism when used independently. Photonic phase-change memory (PCM), which integrates photonics with PCM, overcomes these constraints by enabling simultaneous processing of multiple inputs encoded on different wavelengths, significantly enhancing parallel computation for deep neural network (DNN) inference and training. This paper presents MERIT, a sustainable DNN accelerator that capitalizes on the non-volatility of resistive memory and the high operating speed of photonic devices. MERIT enables seamless inference and training by loading weight kernels into photonic PCM arrays and selectively supplying light encoded with input features for the forward pass and loss gradients for the backward pass. We compare MERIT with state-of-the-art digital and analog DNN accelerators including TPU, DEAP, and PTC. Simulation results demonstrate that MERIT reduces execution time by 68% and energy consumption by 64% for inference, and reduces execution time by 79% and energy consumption by 84% for training. 
    more » « less
  2. ; ; ; (, ACM Transactions on Embedded Computing Systems)
    Domain specific neural network accelerators have garnered attention because of their improved energy efficiency and inference performance compared to CPUs and GPUs. Such accelerators are thus well suited for resource-constrained embedded systems. However, mapping sophisticated neural network models on these accelerators still entails significant energy and memory consumption, along with high inference time overhead. Binarized neural networks (BNNs), which utilize single-bit weights, represent an efficient way to implement and deploy neural network models on accelerators. In this paper, we present a novel optical-domain BNN accelerator, named ROBIN , which intelligently integrates heterogeneous microring resonator optical devices with complementary capabilities to efficiently implement the key functionalities in BNNs. We perform detailed fabrication-process variation analyses at the optical device level, explore efficient corrective tuning for these devices, and integrate circuit-level optimization to counter thermal variations. As a result, our proposed ROBIN architecture possesses the desirable traits of being robust, energy-efficient, low latency, and high throughput, when executing BNN models. Our analysis shows that ROBIN can outperform the best-known optical BNN accelerators and many electronic accelerators. Specifically, our energy-efficient ROBIN design exhibits energy-per-bit values that are ∼4 × lower than electronic BNN accelerators and ∼933 × lower than a recently proposed photonic BNN accelerator, while a performance-efficient ROBIN design shows ∼3 × and ∼25 × better performance than electronic and photonic BNN accelerators, respectively. 
    more » « less
  3. ; ; ; ; ; ; ; ; ; (, Advanced Optical Materials)
    Abstract Neuromorphic photonics has become one of the research forefronts in photonics, with its benefits in low‐latency signal processing and potential in significant energy consumption reduction when compared with digital electronics. With artificial intelligence (AI) computing accelerators in high demand, one of the high‐impact research goals is to build scalable neuromorphic photonic integrated circuits which can accelerate the computing of AI models at high energy efficiency. A complete neuromorphic photonic computing system comprises seven stacks: materials, devices, circuits, microarchitecture, system architecture, algorithms, and applications. Here, we consider microring resonator (MRR)‐based network designs toward building scalable silicon integrated photonic neural networks (PNN), and variations of MRR resonance wavelength from the fabrication process and their impact on PNN scalability. Further, post‐fabrication processing using organic photochromic layers over the silicon platform is shown to be effective for trimming MRR resonance wavelength variation, which can significantly reduce energy consumption from the MRR‐based PNN configuration. Post‐fabrication processing with photochromic materials to compensate for the variation in MRR fabrication will allow a scalable silicon system on a chip without sacrificing today's performance metrics, which will be critical for the commercial viability and volume production of large‐scale silicon photonic circuits. 
    more » « less
  4. ; ; ; (, Optica Publishing Group)
    We propose a method of performing variational inference using high speed photonic neural accelerators. This method incurs no slowdown compared to deterministic photonic inference, affecting only the power consumption of existing accelerator architectures. 
    more » « less
  5. ; ; ; ; ; (, Design, Automation & Test in Europe Conference & Exhibition)
    Although Convolutional Neural Networks (CNNs) have demonstrated the state-of-the-art inference accuracy in various intelligent applications, each CNN inference involves millions of expensive floating point multiply-accumulate (MAC) operations. To energy-efficiently process CNN inferences, prior work proposes an electro-optical accelerator to process power-of-2 quantized CNNs by electro-optical ripple-carry adders and optical binary shifters. The electro-optical accelerator also uses SRAM registers to store intermediate data. However, electro-optical ripple-carry adders and SRAMs seriously limit the operating frequency and inference throughput of the electro-optical accelerator, due to the long critical path of the adder and the long access latency of SRAMs. In this paper, we propose a photonic nonvolatile memory (NVM)-based accelerator, Light-Bulb, to process binarized CNNs by high frequency photonic XNOR gates and popcount units. LightBulb also adopts photonic racetrack memory to serve as input/output registers to achieve high operating frequency. Compared to prior electro-optical accelerators, on average, LightBulb improves the CNN inference throughput by 17× ~ 173× and the inference throughput per Watt by 17.5 × ~ 660×. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email