Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip. Yao, M., Richter, O., Zhao, G., Qiao, N., Xing, Y., Wang, D., Hu, T., Fang, W., Demirci, T., De Marchi, M., Deng, L., Yan, T., Nielsen, C., Sheik, S., Wu, C., Tian, Y., Xu, B., & Li, G. Nature Communications, 15(1):4464, May, 2024. Publisher: Nature Publishing Group![Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex By mimicking the neurons and synapses of the human brain and employing spiking neural networks on neuromorphic chips, neuromorphic computing offers a promising energy-efficient machine intelligence. How to borrow high-level brain dynamic mechanisms to help neuromorphic computing achieve energy advantages is a fundamental issue. This work presents an application-oriented algorithm-software-hardware co-designed neuromorphic system for this issue. First, we design and fabricate an asynchronous chip called “Speck”, a sensing-computing neuromorphic system on chip. With the low processor resting power of 0.42mW, Speck can satisfy the hardware requirements of dynamic computing: no-input consumes no energy. Second, we uncover the “dynamic imbalance” in spiking neural networks and develop an attention-based framework for achieving the algorithmic requirements of dynamic computing: varied inputs consume energy with large variance. Together, we demonstrate a neuromorphic system with real-time power as low as 0.70mW. This work exhibits the promising potentials of neuromorphic computing with its asynchronous event-driven, sparse, and dynamic nature.
@article{yao_spike-based_2024,
title = {Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip},
volume = {15},
copyright = {2024 The Author(s)},
issn = {2041-1723},
url = {https://www.nature.com/articles/s41467-024-47811-6},
doi = {10.1038/s41467-024-47811-6},
abstract = {By mimicking the neurons and synapses of the human brain and employing spiking neural networks on neuromorphic chips, neuromorphic computing offers a promising energy-efficient machine intelligence. How to borrow high-level brain dynamic mechanisms to help neuromorphic computing achieve energy advantages is a fundamental issue. This work presents an application-oriented algorithm-software-hardware co-designed neuromorphic system for this issue. First, we design and fabricate an asynchronous chip called “Speck”, a sensing-computing neuromorphic system on chip. With the low processor resting power of 0.42mW, Speck can satisfy the hardware requirements of dynamic computing: no-input consumes no energy. Second, we uncover the “dynamic imbalance” in spiking neural networks and develop an attention-based framework for achieving the algorithmic requirements of dynamic computing: varied inputs consume energy with large variance. Together, we demonstrate a neuromorphic system with real-time power as low as 0.70mW. This work exhibits the promising potentials of neuromorphic computing with its asynchronous event-driven, sparse, and dynamic nature.},
language = {en},
number = {1},
urldate = {2024-08-21},
journal = {Nature Communications},
author = {Yao, Man and Richter, Ole and Zhao, Guangshe and Qiao, Ning and Xing, Yannan and Wang, Dingheng and Hu, Tianxiang and Fang, Wei and Demirci, Tugba and De Marchi, Michele and Deng, Lei and Yan, Tianyi and Nielsen, Carsten and Sheik, Sadique and Wu, Chenxi and Tian, Yonghong and Xu, Bo and Li, Guoqi},
month = may,
year = {2024},
note = {Publisher: Nature Publishing Group},
keywords = {Computational science, Computer science},
pages = {4464},
}
Downloads: 0
{"_id":"P54hSfrLM9xTpvcKX","bibbaseid":"yao-richter-zhao-qiao-xing-wang-hu-fang-etal-spikebaseddynamiccomputingwithasynchronoussensingcomputingneuromorphicchip-2024","author_short":["Yao, M.","Richter, O.","Zhao, G.","Qiao, N.","Xing, Y.","Wang, D.","Hu, T.","Fang, W.","Demirci, T.","De Marchi, M.","Deng, L.","Yan, T.","Nielsen, C.","Sheik, S.","Wu, C.","Tian, Y.","Xu, B.","Li, G."],"bibdata":{"bibtype":"article","type":"article","title":"Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip","volume":"15","copyright":"2024 The Author(s)","issn":"2041-1723","url":"https://www.nature.com/articles/s41467-024-47811-6","doi":"10.1038/s41467-024-47811-6","abstract":"By mimicking the neurons and synapses of the human brain and employing spiking neural networks on neuromorphic chips, neuromorphic computing offers a promising energy-efficient machine intelligence. How to borrow high-level brain dynamic mechanisms to help neuromorphic computing achieve energy advantages is a fundamental issue. This work presents an application-oriented algorithm-software-hardware co-designed neuromorphic system for this issue. First, we design and fabricate an asynchronous chip called “Speck”, a sensing-computing neuromorphic system on chip. With the low processor resting power of 0.42mW, Speck can satisfy the hardware requirements of dynamic computing: no-input consumes no energy. Second, we uncover the “dynamic imbalance” in spiking neural networks and develop an attention-based framework for achieving the algorithmic requirements of dynamic computing: varied inputs consume energy with large variance. Together, we demonstrate a neuromorphic system with real-time power as low as 0.70mW. This work exhibits the promising potentials of neuromorphic computing with its asynchronous event-driven, sparse, and dynamic nature.","language":"en","number":"1","urldate":"2024-08-21","journal":"Nature Communications","author":[{"propositions":[],"lastnames":["Yao"],"firstnames":["Man"],"suffixes":[]},{"propositions":[],"lastnames":["Richter"],"firstnames":["Ole"],"suffixes":[]},{"propositions":[],"lastnames":["Zhao"],"firstnames":["Guangshe"],"suffixes":[]},{"propositions":[],"lastnames":["Qiao"],"firstnames":["Ning"],"suffixes":[]},{"propositions":[],"lastnames":["Xing"],"firstnames":["Yannan"],"suffixes":[]},{"propositions":[],"lastnames":["Wang"],"firstnames":["Dingheng"],"suffixes":[]},{"propositions":[],"lastnames":["Hu"],"firstnames":["Tianxiang"],"suffixes":[]},{"propositions":[],"lastnames":["Fang"],"firstnames":["Wei"],"suffixes":[]},{"propositions":[],"lastnames":["Demirci"],"firstnames":["Tugba"],"suffixes":[]},{"propositions":[],"lastnames":["De","Marchi"],"firstnames":["Michele"],"suffixes":[]},{"propositions":[],"lastnames":["Deng"],"firstnames":["Lei"],"suffixes":[]},{"propositions":[],"lastnames":["Yan"],"firstnames":["Tianyi"],"suffixes":[]},{"propositions":[],"lastnames":["Nielsen"],"firstnames":["Carsten"],"suffixes":[]},{"propositions":[],"lastnames":["Sheik"],"firstnames":["Sadique"],"suffixes":[]},{"propositions":[],"lastnames":["Wu"],"firstnames":["Chenxi"],"suffixes":[]},{"propositions":[],"lastnames":["Tian"],"firstnames":["Yonghong"],"suffixes":[]},{"propositions":[],"lastnames":["Xu"],"firstnames":["Bo"],"suffixes":[]},{"propositions":[],"lastnames":["Li"],"firstnames":["Guoqi"],"suffixes":[]}],"month":"May","year":"2024","note":"Publisher: Nature Publishing Group","keywords":"Computational science, Computer science","pages":"4464","bibtex":"@article{yao_spike-based_2024,\n\ttitle = {Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip},\n\tvolume = {15},\n\tcopyright = {2024 The Author(s)},\n\tissn = {2041-1723},\n\turl = {https://www.nature.com/articles/s41467-024-47811-6},\n\tdoi = {10.1038/s41467-024-47811-6},\n\tabstract = {By mimicking the neurons and synapses of the human brain and employing spiking neural networks on neuromorphic chips, neuromorphic computing offers a promising energy-efficient machine intelligence. How to borrow high-level brain dynamic mechanisms to help neuromorphic computing achieve energy advantages is a fundamental issue. This work presents an application-oriented algorithm-software-hardware co-designed neuromorphic system for this issue. First, we design and fabricate an asynchronous chip called “Speck”, a sensing-computing neuromorphic system on chip. With the low processor resting power of 0.42mW, Speck can satisfy the hardware requirements of dynamic computing: no-input consumes no energy. Second, we uncover the “dynamic imbalance” in spiking neural networks and develop an attention-based framework for achieving the algorithmic requirements of dynamic computing: varied inputs consume energy with large variance. Together, we demonstrate a neuromorphic system with real-time power as low as 0.70mW. This work exhibits the promising potentials of neuromorphic computing with its asynchronous event-driven, sparse, and dynamic nature.},\n\tlanguage = {en},\n\tnumber = {1},\n\turldate = {2024-08-21},\n\tjournal = {Nature Communications},\n\tauthor = {Yao, Man and Richter, Ole and Zhao, Guangshe and Qiao, Ning and Xing, Yannan and Wang, Dingheng and Hu, Tianxiang and Fang, Wei and Demirci, Tugba and De Marchi, Michele and Deng, Lei and Yan, Tianyi and Nielsen, Carsten and Sheik, Sadique and Wu, Chenxi and Tian, Yonghong and Xu, Bo and Li, Guoqi},\n\tmonth = may,\n\tyear = {2024},\n\tnote = {Publisher: Nature Publishing Group},\n\tkeywords = {Computational science, Computer science},\n\tpages = {4464},\n}\n\n\n\n","author_short":["Yao, M.","Richter, O.","Zhao, G.","Qiao, N.","Xing, Y.","Wang, D.","Hu, T.","Fang, W.","Demirci, T.","De Marchi, M.","Deng, L.","Yan, T.","Nielsen, C.","Sheik, S.","Wu, C.","Tian, Y.","Xu, B.","Li, G."],"key":"yao_spike-based_2024","id":"yao_spike-based_2024","bibbaseid":"yao-richter-zhao-qiao-xing-wang-hu-fang-etal-spikebaseddynamiccomputingwithasynchronoussensingcomputingneuromorphicchip-2024","role":"author","urls":{"Paper":"https://www.nature.com/articles/s41467-024-47811-6"},"keyword":["Computational science","Computer science"],"metadata":{"authorlinks":{}},"downloads":0,"html":""},"bibtype":"article","biburl":"https://bibbase.org/zotero/qiuyuanwang","dataSources":["wWPhSRj9hrZuqsm9D"],"keywords":["computational science","computer science"],"search_terms":["spike","based","dynamic","computing","asynchronous","sensing","computing","neuromorphic","chip","yao","richter","zhao","qiao","xing","wang","hu","fang","demirci","de marchi","deng","yan","nielsen","sheik","wu","tian","xu","li"],"title":"Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip","year":2024}