Zobrazeno 1 - 10
of 107
pro vyhledávání: '"Nahmias, Mitchell"'
Analog electronic and optical computing exhibit tremendous advantages over digital computing for accelerating deep learning when operations are executed at low precision. In this work, we derive a relationship between analog precision, which is limit
Externí odkaz:
http://arxiv.org/abs/2102.06365
Many neural network quantization techniques have been developed to decrease the computational and memory footprint of deep learning. However, these methods are evaluated subject to confounding tradeoffs that may affect inference acceleration or resou
Externí odkaz:
http://arxiv.org/abs/2102.06366
Autor:
Nahmias, Mitchell A., Peng, Hsuan-Tung, de Lima, Thomas Ferreira, Huang, Chaoran, Tait, Alexander N., Shastri, Bhavin J., Prucnal, Paul R.
There has been a recent surge of interest in the implementation of linear operations such as matrix multipications using photonic integrated circuit technology. However, these approaches require an efficient and flexible way to perform nonlinear oper
Externí odkaz:
http://arxiv.org/abs/2012.08516
Autor:
de Lima, Thomas Ferreira, Tait, Alexander N., Saeidi, Hooman, Nahmias, Mitchell A., Peng, Hsuan-Tung, Abbaslou, Siamak, Shastri, Bhavin J., Prucnal, Paul R.
Neuromorphic photonics relies on efficiently emulating analog neural networks at high speeds. Prior work showed that transducing signals from the optical to the electrical domain and back with transimpedance gain was an efficient approach to implemen
Externí odkaz:
http://arxiv.org/abs/1907.07325
Autor:
Bangari, Viraj, Marquez, Bicky A., Miller, Heidi B., Tait, Alexander N., Nahmias, Mitchell A., de Lima, Thomas Ferreira, Peng, Hsuan-Tung, Prucnal, Paul R., Shastri, Bhavin J.
Convolutional Neural Networks (CNNs) are powerful and highly ubiquitous tools for extracting features from large datasets for applications such as computer vision and natural language processing. However, a convolution is a computationally expensive
Externí odkaz:
http://arxiv.org/abs/1907.01525
Autor:
Tait, Alexander N., Ma, Philip Y., de Lima, Thomas Ferreira, Blow, Eric C., Chang, Matthew P., Nahmias, Mitchell A., Shastri, Bhavin J., Prucnal, Paul R.
Multi-antenna radio front-ends generate a multi-dimensional flood of information, most of which is partially redundant. Redundancy is eliminated by dimensionality reduction, but contemporary digital processing techniques face harsh fundamental tradeo
Externí odkaz:
http://arxiv.org/abs/1903.03474
Autor:
Tait, Alexander N., de Lima, Thomas Ferreira, Nahmias, Mitchell A., Miller, Heidi B., Peng, Hsuan-Tung, Shastri, Bhavin J., Prucnal, Paul R.
Publikováno v:
Phys. Rev. Applied 11, 064043 (2019)
There has been a recently renewed interest in neuromorphic photonics, a field promising to access pivotal and unexplored regimes of machine intelligence. Progress has been made on isolated neurons and analog interconnects; nevertheless, this renewal
Externí odkaz:
http://arxiv.org/abs/1812.11898
Autor:
Shastri, Bhavin J., Tait, Alexander N., de Lima, Thomas Ferreira, Nahmias, Mitchell A., Peng, Hsuan-Tung, Prucnal, Paul R.
In an age overrun with information, the ability to process reams of data has become crucial. The demand for data will continue to grow as smart gadgets multiply and become increasingly integrated into our daily lives. Next-generation industries in ar
Externí odkaz:
http://arxiv.org/abs/1801.00016
Autor:
Tait, Alexander N., de Lima, Thomas Ferreira, Zhou, Ellen, Wu, Allie X., Nahmias, Mitchell A., Shastri, Bhavin J., Prucnal, Paul R.
Publikováno v:
Sci.Rep. 7 (2017) 7430
Photonic systems for high-performance information processing have attracted renewed interest. Neuromorphic silicon photonics has the potential to integrate processing functions that vastly exceed the capabilities of electronics. We report first obser
Externí odkaz:
http://arxiv.org/abs/1611.02272
Autor:
Shastri, Bhavin J., Nahmias, Mitchell A., Tait, Alexander N., Rodriguez, Alejandro W., Wu, Ben, Prucnal, Paul R.
Publikováno v:
Scientific Reports 6, Article number: 19126 (2016)
Novel materials and devices in photonics have the potential to revolutionize optical information processing, beyond conventional binary-logic approaches. Laser systems offer a rich repertoire of useful dynamical behaviors, including the excitable dyn
Externí odkaz:
http://arxiv.org/abs/1507.06713