Cloud and Edge: Price, Performance and Privacy Considerations in IOT
summary
NEC Labs adapts deep learning to run privately on edge and cloud by using private computing, combining secure enclaves (SGX/trustzone), homomorphic encryption and multi-party computation with TensorFlow/Keras-style pipelines. MPC performance improves from ~10^6× slower historically to about ~10^3× slower today, making banking NLP, speech and many image models practical in pilots (video analytics still trailing but within a year). Deployment model routes secret shares via CSPs/carriers (IOtopia) so devices stay unchanged; code is open source and actively piloted.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 4:00IOT challenges and security risks
- 9:41Regulations and data privacy laws
- 11:54Private computing explained
- 18:23Practical applications of private computing
- 23:20Future of IOT with private computing
- 26:32Call to action and consortium
related talks
36mApplied AI on the edge
49mEnable IoT with Edge Computing and Machine Learning
57mWhy should you care about edge computing?
01h 01mML and the IoT: Living on the Edge
20mHow ML is Powering the Future of IoT
48mThe Power of Edge AI: Revolutionizing Future of Real-Time, Autonomous Processing
59mML and the IoT: Living on the Edge
37mDeep Learning Neural Network Acceleration at the Edge