Freely scalable and reconfigurable optical hardware for deep learning

Abstract

As deep neural network (DNN) models grow ever-larger, they can achieve higher accuracy and solve more complex problems. This trend has been enabled by an increase in available compute power; however, efforts to continue to scale electronic processors are impeded by the costs of communication, thermal management, power delivery and clocking. To improve scalability, we propose a digital optical neural network (DONN) with intralayer optical interconnects and reconfigurable input values. The near path-length-independence of optical energy consumption enables information locality between a transmitter and arbitrarily arranged receivers, which allows greater flexibility in architecture design to circumvent scaling limitations. In a proof-of-concept experiment, we demonstrate optical multicast in the classification of 500 MNIST images with a 3-layer, fully-connected network. We also analyze the energy consumption of the DONN and find that optical data transfer is beneficial over electronics when the spacing of computational units is on the order of >10 µm.

Deep-learning-based quantum imaging using NOON states

“…If so, a more convient way to generate or measure the large-N NOON states can be expected, whose cost would be shrinked to the level of that by generating or measuring the small-N NOON states. Meanwhile, as we briefly discussed in the above, the recent progress in classcial optics [32][33][34]41] have shown that the optical nerual networks exhibit a great potential for light modulation. Due to the Klyshko's advanced-wave picture [42], we believe that those optcal nerual networks are also workable for the manipulation of quantum light.…”

“…This means the method could provide a rather friendly starting point for the daily application of NOON state imaging. Besides, we notice that a series of works on optical implmenations of nerual network have been presented [32][33][34]. These work indicates that well-trained nerual networks can provide new insights about designing optical elements for specific aim.…”

Large-scale Sustainable Search on Unconventional Computing Hardware

“…Nonlinearities are introduced intentionally in other physical computing platforms and can be omit-ted to model the PageRank algorithm. For example, the PageRank can be calculated by performing optical matrix multiplications that support beyond GHz clock rates [40][41][42].…”

“…The various platforms for such optimisation include optical parametric oscillators [17,18], electronic oscillators [19,20], memristors [21], lasers [22][23][24][25], photonic simulators [26,27], cold atoms [28,29], trapped ions [30], polariton condensates [31,32], photon condensates [33], QED [34,35], and others [36][37][38]. While the demonstration of their ability to find the global minima of computationally hard problems faster than the classical von Neumann architecture remains elusive, many of these disparate physical systems can either efficiently perform matrix-vector multiplication [26,[39][40][41][42] or mimic the Hopfield neural networks [21,43,44]. For a certain choice of parameters, the time evolution of such networks can be viewed as an eigenvalue maximisation problem [45], which results in finding the energy state dictated by signs of the eigenvector corresponding to the largest eigenvalue of the interaction matrix, i.e.…”

Robust calibration of multiparameter sensors via machine learning at the single-photon level

“…In the last few years, several applications of machine learning methods in the quantum domain have been reported [14][15][16], including state and unitary tomography [17][18][19][20][21][22][23][24][25], design of quantum experiments [26][27][28][29][30][31][32], validation of quantum technology [33][34][35], identification of quantum features [36,37], and the adaptive control of quantum devices [38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53][54]. Also, photonic platforms can be exploited for the realization of machine learning protocols [55,56]. Recently, a first insight on the application of machine learning methods for the calibration of a quantum sensor has been reported [57].…”