We are excited to bring Transform 2022 back to life on 19th July and virtually 20th July – 3rd August. Join AI and data leaders for sensible conversations and exciting networking opportunities. Learn more about Transform 2022
CEA-Leti, a French research institute, has received a તરફથી 3 million grant from the European Research Council (ERC) to create the first type of intelligent edge AI system inspired by the insect’s nervous system. The Nanoscale system uses innovative technologies to enhance the performance and energy-efficiency of Edge AI. Targeted applications include robotics, such as fruit picking and rescue, medical implants and wearable electronics.
Although classical computing architectures used in electronics such as PCs, data centers and mobile devices are ubiquitous today, one problem is that most of the energy is actually used to move data instead of processing data. That’s why over time researchers have tried to find better approaches, especially given the rise of AI in the last decade. One such option is called in-memory computing, where memory is used for both storage and processing.
Bugged inspiration for edge AI
However, the problem with in-memory computing is that it requires fast, non-volatile memory with high endurance. Such memory does not currently exist because DRAM is unstable. To overcome this problem and reduce the need for memory, CEA-Leti scientists have been inspired by the insect’s nervous system.
Senior scientist Alyssa Vianello has received a 3 million grant from ERC to use new nanoscale memory technology that mimics biological mechanisms from insects to create silicon-based, energy-efficient nanoscale systems for the same AI.
“My project is to draw inspiration from insects’ nervous systems to alleviate hardware requirements in terms of memory density and reliability, and to create new nanosystems that we need to be able to learn from very limited volumes of noisy data,” Vianello said.
“Cricketers make accurate decisions based on sluggish, inaccurate and unreliable neurons and synapses to avoid their predators. A closer look at their biology reveals a variety of memory-like functions in their sensory and nervous systems. Combining these different functions, Cricket’s internal computing system achieves incredible performance and energy efficiency.
The primary goal is to create devices that enable learning from a limited amount of noisy data. Such data can come from sensors such as video cameras, radar, ECG, EMG, bio-impedance streams and brain signals. To that end, Vianello discovered that the various functions of the insect’s nervous system closely resemble those of the various memory technologies employed by CEA-Letty. As such, scientists aim to create “hybrid synapses” that co-integrate these different memory techniques.
The concept of heterogeneous distributed computing system refers to various processing units and sensors connected by spiking and non-spiking links.
Venturebeat’s mission Digital Town Square is set to become a place for technical decision makers to gain knowledge about the changing enterprise technology and practices. Learn more about membership.