A Coupled CMOS Oscillator Array for 8ns and 55pJ Inference in Convolutional Neural Networks

التفاصيل البيبلوغرافية
العنوان: A Coupled CMOS Oscillator Array for 8ns and 55pJ Inference in Convolutional Neural Networks
المؤلفون: Nikonov, D. E., Kurahashi, P., Ayers, J. S., Lee, H. -J., Fan, Y., Young, I. A.
سنة النشر: 2019
المجموعة: Computer Science
Condensed Matter
Physics (Other)
مصطلحات موضوعية: Computer Science - Emerging Technologies, Condensed Matter - Disordered Systems and Neural Networks, Physics - Applied Physics
الوصف: Oscillator neural networks (ONN) based on arrays of 26 CMOS ring oscillators designed and fabricated. ONN are used for inference of dot products with image fragments and kernels necessary for convolutional neural networks. The inputs are encoded as frequency shifts of oscillators using current DACs. Degree of match (DOM) is determined from oscillators synchronization. Measurements demonstrate high correlation of DOM and dot products. Inference requires the time of 8ns and energy of 55pJ.
Comment: 7 pages, 7 figures
نوع الوثيقة: Working Paper
URL الوصول: http://arxiv.org/abs/1910.11803
رقم الأكسشن: edsarx.1910.11803
قاعدة البيانات: arXiv