Neuromorphic Photonics Based on Phase Change Materials

Abstract

Neuromorphic photonics devices based on phase change materials (PCMs) and silicon photonics technology have emerged as promising solutions for addressing the limitations of traditional spiking neural networks in terms of scalability, response delay, and energy consumption. In this review, we provide a comprehensive analysis of various PCMs used in neuromorphic devices, comparing their optical properties and discussing their applications. We explore materials such as GST (Ge2Sb2Te5), GeTe-Sb2Te3, GSST (Ge2Sb2Se4Te1), Sb2S3/Sb2Se3, Sc0.2Sb2Te3 (SST), and In2Se3, highlighting their advantages and challenges in terms of erasure power consumption, response rate, material lifetime, and on-chip insertion loss. By investigating the integration of different PCMs with silicon-based optoelectronics, this review aims to identify potential breakthroughs in computational performance and scalability of photonic spiking neural networks. Further research and development are essential to optimize these materials and overcome their limitations, paving the way for more efficient and high-performance photonic neuromorphic devices in artificial intelligence and high-performance computing applications.

Magnetization reversal by multiple optical pulses for a photonic spiking neuron with the leaky integrate and fire model

“…In terms of temperature, our structure needed about 490 K which is much less than around 877 K that is required for PCMs to transition from crystal state to amorphous state. 27,29,36) However, an array of ∼10 cells of MO synapse is needed to obtain sufficient phase shift along MO material. As a result, the order of required pulse energy for our MO synapse is almost the same as that of PCM-based synapses.…”

“…20) Utilizing PCMs is one of the most well-established spiking neurons which can make a binary state by transiting back and forth between the state of amorphous and crystal. [27][28][29] Although these studies achieved promising results in the field of photonic SNNs, PCM has challenges such as a rewritable limit and limited phase transitions between amorphous and crystal states because of its degradation due to repeat operation. Therefore, photonic SNNs with high rewritability and robustness against material degradation are desired.…”

Neuromorphic Photonics Circuits: Contemporary Review

“…This paves the way for creating high-speed and energy-frugal computing systems, markedly advantageous for tasks like pattern recognition, machine learning, and intricate data analysis [109]. Moreover, the realm of neuromorphic photonics extends its reach into optical neural networks, poised to reshape data processing across various domains, including telecommunications, image recognition, and autonomous vehicles [110][111][112]. Additionally, it plays a pivotal role in elevating the performance of brain-machine interfaces, facilitating smoother interactions between humans and computers.…”

Editorial of the Special Issue ‘Nano-Optics and Nano-Optoelectronics: Challenges and Future Trends’

“…This Special Issue also encompasses the latest advancements in practical nano-optical and nano-optoelectronic technologies, such as femtosecond laser nano-processing [ 12 ], atomic layer coating [ 13 ], and new designs for photonic integration engineering [ 14 ]. It concludes with reviews that aim to offer comprehensive overviews of nano-related infrared absorption technology [ 15 ] and neuromorphic photonics devices [ 16 ], and to predict the futural trends in these nano-optic and nano-optoelectronic topics. In the following paragraphs, a concise summary of each article will be presented, to pique the interest of potential readers.…”

“…They highlight the emerging needs that have driven the development of compact and integrated infrared spectroscopy systems and chips, while also emphasizing the role of machine learning in facilitating device design and data analysis. Li et al [ 16 ] present a comprehensive review of neuromorphic photonic devices that make use of phase-change materials (PCMs) and silicon photonic technology. They conduct a comprehensive analysis of various PCMs employed in neuromorphic devices, comparing their optical properties and discussing their applications.…”