SamurAI: A Versatile IoT Node With Event-Driven Wake-Up and Embedded ML Acceleration
Authors:
Ivan Miro-Panades,
Benoit Tain,
Jean-Frederic Christmann,
David Coriat,
Romain Lemaire,
Clement Jany,
Baudouin Martineau,
Fabrice Chaix,
Guillaume Waltener,
Emmanuel Pluchart,
Jean-Philippe Noel,
Adam Makosiej,
Maxime Montoya,
Simone Bacles-Min,
David Briand,
Jean-Marc Philippe,
Yvain Thonnart,
Alexandre Valentian,
Frederic Heitzmann,
Fabien Clermidy
Abstract:
Increased capabilities such as recognition and self-adaptability are now required from IoT applications. While IoT node power consumption is a major concern for these applications, cloud-based processing is becoming unsustainable due to continuous sensor or image data transmission over the wireless network. Thus optimized ML capabilities and data transfers should be integrated in the IoT node. Mor…
▽ More
Increased capabilities such as recognition and self-adaptability are now required from IoT applications. While IoT node power consumption is a major concern for these applications, cloud-based processing is becoming unsustainable due to continuous sensor or image data transmission over the wireless network. Thus optimized ML capabilities and data transfers should be integrated in the IoT node. Moreover, IoT applications are torn between sporadic data-logging and energy-hungry data processing (e.g. image classification). Thus, the versatility of the node is key in addressing this wide diversity of energy and processing needs. This paper presents SamurAI, a versatile IoT node bridging this gap in processing and in energy by leveraging two on-chip sub-systems: a low power, clock-less, event-driven Always-Responsive (AR) part and an energy-efficient On-Demand (OD) part. AR contains a 1.7MOPS event-driven, asynchronous Wake-up Controller (WuC) with a 207ns wake-up time optimized for sporadic computing, while OD combines a deep-sleep RISC-V CPU and 1.3TOPS/W Machine Learning (ML) for more complex tasks up to 36GOPS. This architecture partitioning achieves best in class versatility metrics such as peak performance to idle power ratio. On an applicative classification scenario, it demonstrates system power gains, up to 3.5x compared to cloud-based processing, and thus extended battery lifetime.
△ Less
Submitted 11 April, 2023;
originally announced April 2023.
Density-based clustering of crystal orientations and misorientations and the orix python library
Authors:
Duncan N. Johnstone,
Ben H. Martineau,
Phillip Crout,
Paul A. Midgley,
Alexander S. Eggeman
Abstract:
Crystal orientation mapping experiments typically measure orientations that are similar within grains and misorientations that are similar along grain boundaries. Such (mis)orientation data will cluster in (mis)orientation space and clusters are more pronounced if preferred orientations or special orientation relationships are present. Here, cluster analysis of (mis)orientation data is described a…
▽ More
Crystal orientation mapping experiments typically measure orientations that are similar within grains and misorientations that are similar along grain boundaries. Such (mis)orientation data will cluster in (mis)orientation space and clusters are more pronounced if preferred orientations or special orientation relationships are present. Here, cluster analysis of (mis)orientation data is described and demonstrated using distance metrics incorporating crystal symmetry and the density-based clustering algorithm DBSCAN. Frequently measured (mis)orientations are identified as corresponding to grains, grain boundaries or orientation relationships, which are visualised both spatially and in three-dimensional (mis)orientation spaces. A new open-source python library, orix, is also reported.
△ Less
Submitted 8 January, 2020;
originally announced January 2020.
