Pub Date : 2024-06-06DOI: 10.1038/s44287-024-00050-9
Nikolaos Farmakidis, Bowei Dong, Harish Bhaskaran
Using photons in lieu of electrons to process information has been an exciting technological prospect for decades. Optical computing is gaining renewed enthusiasm, owing to the accumulated maturity of photonic integrated circuits and the pressing need for faster processing to cope with data generated by artificial intelligence. In neuromorphic photonics, the bosonic nature of light is exploited for high-speed, densely multiplexed linear operations, whereas the superior computing modalities of biological neurons are imitated to accelerate computations. Here, we provide an overview of recent advances in integrated synaptic optical devices and on-chip photonic neural networks focusing on the location in the architecture at which the optical to electrical conversion takes place. We present challenges associated with electro-optical conversions, implementations of optical nonlinearity, amplification and processing in the time domain, and we identify promising emerging photonic neuromorphic hardware. Neuromorphic photonics is an emerging computing platform that addresses the growing computational demands of modern society. We review advances in integrated neuromorphic photonics and discuss challenges associated with electro-optical conversions, implementations of nonlinearity, amplification and processing in the time domain.
{"title":"Integrated photonic neuromorphic computing: opportunities and challenges","authors":"Nikolaos Farmakidis, Bowei Dong, Harish Bhaskaran","doi":"10.1038/s44287-024-00050-9","DOIUrl":"10.1038/s44287-024-00050-9","url":null,"abstract":"Using photons in lieu of electrons to process information has been an exciting technological prospect for decades. Optical computing is gaining renewed enthusiasm, owing to the accumulated maturity of photonic integrated circuits and the pressing need for faster processing to cope with data generated by artificial intelligence. In neuromorphic photonics, the bosonic nature of light is exploited for high-speed, densely multiplexed linear operations, whereas the superior computing modalities of biological neurons are imitated to accelerate computations. Here, we provide an overview of recent advances in integrated synaptic optical devices and on-chip photonic neural networks focusing on the location in the architecture at which the optical to electrical conversion takes place. We present challenges associated with electro-optical conversions, implementations of optical nonlinearity, amplification and processing in the time domain, and we identify promising emerging photonic neuromorphic hardware. Neuromorphic photonics is an emerging computing platform that addresses the growing computational demands of modern society. We review advances in integrated neuromorphic photonics and discuss challenges associated with electro-optical conversions, implementations of nonlinearity, amplification and processing in the time domain.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 6","pages":"358-373"},"PeriodicalIF":0.0,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141381882","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Driven by the global effort towards reduction of carbon dioxide emissions from cars, the gradual phase out of fuel cars accompanied by the rise of electric vehicles (EVs) has become a megatrend. Despite the rapid growth of electric vehicle markets worldwide, the leading manufacturers recently announced notable price reductions to compete for market shares. From the technology perspective, for fast charging and extended driving range, more electric vehicles now shift to 800-V batteries with the traction inverters based on wide-bandgap SiC, which can lead to higher efficiency and higher power densities compared with the Si counterparts. However, to further reduce the SiC substrate and epitaxy cost remains a challenge. By contrast, for the DC–DC converters and onboard chargers of electric vehicles, the power switches based on GaN enable fast switching, which can significantly reduce the module form factors. However, the high-voltage reliability concerns associated with the heteroepitaxial defects affect the widespread adoption of GaN in electric vehicles. In this Review, we present a comprehensive discussion of the state-of-the-art power electronics for electric vehicles based on Si, SiC and GaN technologies from the device to circuit and module levels. Various competing technologies are evaluated in consideration of not only efficiency but also cost and reliability, which constitute the three main pillars supporting the continuous growth of electric vehicle power electronics. This Review discusses the state-of-the-art power electronics in electric vehicles based on Si, SiC and GaN from an industry perspective, with a particular focus on the module power densities, efficiencies, costs and reliabilities with the 800-V battery.
在全球努力减少汽车二氧化碳排放的推动下，燃油汽车的逐步淘汰和电动汽车（EV）的兴起已成为一个大趋势。尽管全球电动汽车市场增长迅速，但主要制造商最近都宣布大幅降价，以争夺市场份额。从技术角度来看，为了实现快速充电和延长行驶里程，目前越来越多的电动汽车转向使用 800-V 电池和基于宽带隙碳化硅的牵引逆变器，这与碳化硅相比，可以实现更高的效率和功率密度。然而，进一步降低碳化硅衬底和外延成本仍是一项挑战。相比之下，对于电动汽车的直流-直流转换器和车载充电器，基于氮化镓的功率开关可实现快速开关，从而大幅降低模块外形尺寸。然而，与异外延缺陷相关的高压可靠性问题影响了氮化镓在电动汽车中的广泛应用。在本综述中，我们全面讨论了基于硅、碳化硅和氮化镓技术的最先进的电动汽车电力电子技术，从器件到电路和模块层面。在评估各种竞争技术时，我们不仅考虑了效率，还考虑了成本和可靠性，而这正是支撑电动汽车电力电子技术持续发展的三大支柱。本综述从行业角度讨论了基于硅、碳化硅和氮化镓的最先进的电动汽车电力电子技术，尤其关注 800 V 电池的模块功率密度、效率、成本和可靠性。
{"title":"Industry perspective on power electronics for electric vehicles","authors":"Chang-Ching Tu, Chia-Lung Hung, Kuo-Bin Hong, Surya Elangovan, Wei-Chen Yu, Yu-Sheng Hsiao, Wei-Cheng Lin, Rustam Kumar, Zhen-Hong Huang, Yu-Heng Hong, Yi-Kai Hsiao, Ray-Hua Horng, Bing-Yue Tsui, Tian-Li Wu, Jr-Hau He, Hao-Chung Kuo","doi":"10.1038/s44287-024-00055-4","DOIUrl":"10.1038/s44287-024-00055-4","url":null,"abstract":"Driven by the global effort towards reduction of carbon dioxide emissions from cars, the gradual phase out of fuel cars accompanied by the rise of electric vehicles (EVs) has become a megatrend. Despite the rapid growth of electric vehicle markets worldwide, the leading manufacturers recently announced notable price reductions to compete for market shares. From the technology perspective, for fast charging and extended driving range, more electric vehicles now shift to 800-V batteries with the traction inverters based on wide-bandgap SiC, which can lead to higher efficiency and higher power densities compared with the Si counterparts. However, to further reduce the SiC substrate and epitaxy cost remains a challenge. By contrast, for the DC–DC converters and onboard chargers of electric vehicles, the power switches based on GaN enable fast switching, which can significantly reduce the module form factors. However, the high-voltage reliability concerns associated with the heteroepitaxial defects affect the widespread adoption of GaN in electric vehicles. In this Review, we present a comprehensive discussion of the state-of-the-art power electronics for electric vehicles based on Si, SiC and GaN technologies from the device to circuit and module levels. Various competing technologies are evaluated in consideration of not only efficiency but also cost and reliability, which constitute the three main pillars supporting the continuous growth of electric vehicle power electronics. This Review discusses the state-of-the-art power electronics in electric vehicles based on Si, SiC and GaN from an industry perspective, with a particular focus on the module power densities, efficiencies, costs and reliabilities with the 800-V battery.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 7","pages":"435-452"},"PeriodicalIF":0.0,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141380647","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-04DOI: 10.1038/s44287-024-00065-2
Lishu Wu
An article in IEEE Transactions on Wireless Communications proposes solutions for interference management in vehicle-to-everything communication systems by leveraging a one-layer rate-splitting multiple-access scheme.
IEEE Transactions on Wireless Communications》上的一篇文章提出了利用单层速率分割多路访问方案进行车对车通信系统干扰管理的解决方案。
{"title":"Rate-splitting multiple-access-enabled V2X communications","authors":"Lishu Wu","doi":"10.1038/s44287-024-00065-2","DOIUrl":"10.1038/s44287-024-00065-2","url":null,"abstract":"An article in IEEE Transactions on Wireless Communications proposes solutions for interference management in vehicle-to-everything communication systems by leveraging a one-layer rate-splitting multiple-access scheme.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 6","pages":"357-357"},"PeriodicalIF":0.0,"publicationDate":"2024-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141255117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-29DOI: 10.1038/s44287-024-00062-5
Mark Nikolka, Sebastian Göke, Ondrej Burkacky, Peter Spiller, Mark Patel
Driven by trends such as GenAI, Automation and E-mobility, the global semiconductor demand is surging, consequently increasing the industry’s emissions. Given the increasing pressure for decarbonization — for example, from customers such as Apple, Google and Microsoft — semiconductor players need to increase their decarbonization efforts.
{"title":"Unlocking net-zero in semiconductor manufacturing","authors":"Mark Nikolka, Sebastian Göke, Ondrej Burkacky, Peter Spiller, Mark Patel","doi":"10.1038/s44287-024-00062-5","DOIUrl":"10.1038/s44287-024-00062-5","url":null,"abstract":"Driven by trends such as GenAI, Automation and E-mobility, the global semiconductor demand is surging, consequently increasing the industry’s emissions. Given the increasing pressure for decarbonization — for example, from customers such as Apple, Google and Microsoft — semiconductor players need to increase their decarbonization efforts.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 8","pages":"487-488"},"PeriodicalIF":0.0,"publicationDate":"2024-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141195441","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Flat optical components (metasurfaces) made from artificial electromagnetic materials (metamaterials) have opened new possibilities for the manipulation of electromagnetic waves within compact multifunctional devices. This field encompasses the development of individual optical elements and their integration into systems for use in real-world applications, especially in optical information processing, light detection and ranging (LiDAR), augmented or virtual reality, and biological imaging. This comprehensive Review highlights advances in the use of flat optics in analog computational information processing and imaging applications and emphasizes their fundamental role in transfer function engineering. The natural synergy between flat optics and computational technologies is described in relation to advanced imaging and microscopy solutions that have the potential for simultaneous information acquisition and processing. An outlook on future developments, including critical insights for both newcomers and experts in this field, is also provided. Flat optics enable light manipulation at the subwavelength scale and provide a compact, wave-based, information processing and acquisition platform. Here, Wang et al. focus on the emerging interdisciplinary field of computational flat optic imaging applications and reveal their intrinsic connections.
{"title":"Advances in information processing and biological imaging using flat optics","authors":"Xinwei Wang, Huijie Hao, Xiaoyuan He, Peng Xie, Jian Liu, Jiubin Tan, Haoyu Li, Hao Wang, Patrice Genevet, Yu Luo, Xumin Ding, Guangwei Hu","doi":"10.1038/s44287-024-00057-2","DOIUrl":"10.1038/s44287-024-00057-2","url":null,"abstract":"Flat optical components (metasurfaces) made from artificial electromagnetic materials (metamaterials) have opened new possibilities for the manipulation of electromagnetic waves within compact multifunctional devices. This field encompasses the development of individual optical elements and their integration into systems for use in real-world applications, especially in optical information processing, light detection and ranging (LiDAR), augmented or virtual reality, and biological imaging. This comprehensive Review highlights advances in the use of flat optics in analog computational information processing and imaging applications and emphasizes their fundamental role in transfer function engineering. The natural synergy between flat optics and computational technologies is described in relation to advanced imaging and microscopy solutions that have the potential for simultaneous information acquisition and processing. An outlook on future developments, including critical insights for both newcomers and experts in this field, is also provided. Flat optics enable light manipulation at the subwavelength scale and provide a compact, wave-based, information processing and acquisition platform. Here, Wang et al. focus on the emerging interdisciplinary field of computational flat optic imaging applications and reveal their intrinsic connections.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 6","pages":"391-411"},"PeriodicalIF":0.0,"publicationDate":"2024-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141170235","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-20DOI: 10.1038/s44287-024-00052-7
Olga Krestinskaya, Mohammed E. Fouda, Hadjer Benmeziane, Kaoutar El Maghraoui, Abu Sebastian, Wei D. Lu, Mario Lanza, Hai Li, Fadi Kurdahi, Suhaib A. Fahmy, Ahmed Eltawil, Khaled N. Salama
The rapid growth of artificial intelligence and the increasing complexity of neural network models are driving demand for efficient hardware architectures that can address power-constrained and resource-constrained deployments. In this context, the emergence of in-memory computing (IMC) stands out as a promising technology. For this purpose, several IMC devices, circuits and architectures have been developed. However, the intricate nature of designing, implementing and deploying such architectures necessitates a well-orchestrated toolchain for hardware–software co-design. This toolchain must allow IMC-aware optimizations across the entire stack, encompassing devices, circuits, chips, compilers, software and neural network design. The complexity and sheer size of the design space involved renders manual optimizations impractical. To mitigate these challenges, hardware-aware neural architecture search (HW-NAS) has emerged as a promising approach to accelerate the design of streamlined neural networks tailored for efficient deployment on IMC hardware. This Review illustrates the application of HW-NAS to the specific features of IMC hardware and compares existing optimization frameworks. Ongoing research and unresolved issues are discussed. A roadmap for the evolution of HW-NAS for IMC architectures is proposed. Hardware-aware neural architecture search (HW-NAS) can be used to design efficient in-memory computing (IMC) hardware for deep learning accelerators. This Review discusses methodologies, frameworks, ongoing research, open issues and recommendations, and provides a roadmap for HW-NAS for IMC.
{"title":"Neural architecture search for in-memory computing-based deep learning accelerators","authors":"Olga Krestinskaya, Mohammed E. Fouda, Hadjer Benmeziane, Kaoutar El Maghraoui, Abu Sebastian, Wei D. Lu, Mario Lanza, Hai Li, Fadi Kurdahi, Suhaib A. Fahmy, Ahmed Eltawil, Khaled N. Salama","doi":"10.1038/s44287-024-00052-7","DOIUrl":"10.1038/s44287-024-00052-7","url":null,"abstract":"The rapid growth of artificial intelligence and the increasing complexity of neural network models are driving demand for efficient hardware architectures that can address power-constrained and resource-constrained deployments. In this context, the emergence of in-memory computing (IMC) stands out as a promising technology. For this purpose, several IMC devices, circuits and architectures have been developed. However, the intricate nature of designing, implementing and deploying such architectures necessitates a well-orchestrated toolchain for hardware–software co-design. This toolchain must allow IMC-aware optimizations across the entire stack, encompassing devices, circuits, chips, compilers, software and neural network design. The complexity and sheer size of the design space involved renders manual optimizations impractical. To mitigate these challenges, hardware-aware neural architecture search (HW-NAS) has emerged as a promising approach to accelerate the design of streamlined neural networks tailored for efficient deployment on IMC hardware. This Review illustrates the application of HW-NAS to the specific features of IMC hardware and compares existing optimization frameworks. Ongoing research and unresolved issues are discussed. A roadmap for the evolution of HW-NAS for IMC architectures is proposed. Hardware-aware neural architecture search (HW-NAS) can be used to design efficient in-memory computing (IMC) hardware for deep learning accelerators. This Review discusses methodologies, frameworks, ongoing research, open issues and recommendations, and provides a roadmap for HW-NAS for IMC.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 6","pages":"374-390"},"PeriodicalIF":0.0,"publicationDate":"2024-05-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s44287-024-00052-7.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141121837","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-17DOI: 10.1038/s44287-024-00043-8
Christian Cipriani
Current human–machine interfaces for controlling assistive devices fail to offer direct, arbitrary control over multiple degrees of freedom. Based on the implantation and tracking of small magnets in the residual muscles, the myokinetic interface could enable biomimetic, direct, independent and parallel control of artificial limbs.
{"title":"Towards prosthetic limbs and assistive devices controlled via the myokinetic interface","authors":"Christian Cipriani","doi":"10.1038/s44287-024-00043-8","DOIUrl":"10.1038/s44287-024-00043-8","url":null,"abstract":"Current human–machine interfaces for controlling assistive devices fail to offer direct, arbitrary control over multiple degrees of freedom. Based on the implantation and tracking of small magnets in the residual muscles, the myokinetic interface could enable biomimetic, direct, independent and parallel control of artificial limbs.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 5","pages":"282-283"},"PeriodicalIF":0.0,"publicationDate":"2024-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140962369","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-17DOI: 10.1038/s44287-024-00042-9
Jiaqi Ruan, Zhao Xu, Hui Su
The inherent differences in epistemologies and research methods in electrical engineering and earth science hinder interdisciplinary collaboration. In the context of climate change, this divide affects the shift towards long-term sustainability in global energy systems, prompting dialogue between the disciplines to enable effective interdisciplinary collaborations.
{"title":"Towards interdisciplinary integration of electrical engineering and earth science","authors":"Jiaqi Ruan, Zhao Xu, Hui Su","doi":"10.1038/s44287-024-00042-9","DOIUrl":"10.1038/s44287-024-00042-9","url":null,"abstract":"The inherent differences in epistemologies and research methods in electrical engineering and earth science hinder interdisciplinary collaboration. In the context of climate change, this divide affects the shift towards long-term sustainability in global energy systems, prompting dialogue between the disciplines to enable effective interdisciplinary collaborations.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 5","pages":"278-279"},"PeriodicalIF":0.0,"publicationDate":"2024-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140963713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-17DOI: 10.1038/s44287-024-00060-7
Networking is an essential skill that offers several perks, from improving the visibility of research by effective communication with peers and scientific editors to advancing careers and building a lasting reputation.
{"title":"Growing together through networking","authors":"","doi":"10.1038/s44287-024-00060-7","DOIUrl":"10.1038/s44287-024-00060-7","url":null,"abstract":"Networking is an essential skill that offers several perks, from improving the visibility of research by effective communication with peers and scientific editors to advancing careers and building a lasting reputation.","PeriodicalId":501701,"journal":{"name":"Nature Reviews Electrical Engineering","volume":"1 5","pages":"275-275"},"PeriodicalIF":0.0,"publicationDate":"2024-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s44287-024-00060-7.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141060099","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：