Eco2AI: carbon emissions tracking of machine learning models as the first step towards sustainable AI
Authors:
Semen Budennyy,
Vladimir Lazarev,
Nikita Zakharenko,
Alexey Korovin,
Olga Plosskaya,
Denis Dimitrov,
Vladimir Arkhipkin,
Ivan Oseledets,
Ivan Barsola,
Ilya Egorov,
Aleksandra Kosterina,
Leonid Zhukov
Abstract:
The size and complexity of deep neural networks continue to grow exponentially, significantly increasing energy consumption for training and inference by these models. We introduce an open-source package eco2AI to help data scientists and researchers to track energy consumption and equivalent CO2 emissions of their models in a straightforward way. In eco2AI we put emphasis on accuracy of energy co…
▽ More
The size and complexity of deep neural networks continue to grow exponentially, significantly increasing energy consumption for training and inference by these models. We introduce an open-source package eco2AI to help data scientists and researchers to track energy consumption and equivalent CO2 emissions of their models in a straightforward way. In eco2AI we put emphasis on accuracy of energy consumption tracking and correct regional CO2 emissions accounting. We encourage research community to search for new optimal Artificial Intelligence (AI) architectures with a lower computational cost. The motivation also comes from the concept of AI-based green house gases sequestrating cycle with both Sustainable AI and Green AI pathways.
△ Less
Submitted 3 August, 2022; v1 submitted 31 July, 2022;
originally announced August 2022.
Software-defined subcarrier wave quantum networking operated by OpenFlow protocol
Authors:
Vladimir V. Chistyakov,
Oleg L. Sadov,
Artur B. Vasiliev,
Vladimir I. Egorov,
Mikhail V. Kompaniets,
Pyotr V. Fedchenkov,
Oleg I. Lazo,
Andrey E. Shevel,
Nikita V. Buldakov,
Artur V. Gleim,
Sergei E. Khoruzhnikov
Abstract:
Future practical implementation of secure quantum communication technology in a multiuser network environment would require automatic monitoring of the optical link condition and quantum system parameters, along with adjusting them in real time according to protocol restrictions. We demonstrate how software-defined networking (SDN) paradigm can be used for addressing this problem on the example of…
▽ More
Future practical implementation of secure quantum communication technology in a multiuser network environment would require automatic monitoring of the optical link condition and quantum system parameters, along with adjusting them in real time according to protocol restrictions. We demonstrate how software-defined networking (SDN) paradigm can be used for addressing this problem on the example of a subcarrier wave quantum communication system, which is promising for network applications. We propose a novel approach to dynamic quantum network routing and communication security based on SDN OpenFlow protocol. SDN-operated dynamic switching between different data encryption methods, quantum or classic, will enable to organize virtual encoding channels with an ability to set the required Quality of Service level for security/bandwidth. These developments will further increase the feasibility of quantum technologies from the network perspective and contribute to bringing them to industrial scale.
△ Less
Submitted 26 September, 2017;
originally announced September 2017.
