News

Acceleration Robotics launch ROBOTCORE™ to speed-up ROS 2 robots

Source: businesswire, June 16, 2022 [1]
The ROBOTCORE™ hardware acceleration framework helps roboticists build “robot chips” and make robot computations faster, more deterministic and power efficient through FPGAs and GPUs.

ROS 2 driver now available for ABB’s robot arms

Source: The Robot Report, June 10, 2022 [2]
A ROS 2 driver for ABB manipulators has been released. The package allows two modes of simulation. It can spoof ros2_control drivers for simulation in ROS. Alternatively, it can communicate with ABB's own RobotStudio software.

Microchip pushes first RISC-V-based SoC FPGA to mass production

Source: allaboutcircuits, June 11, 2022 [3]
Microchip continues the push for RISC-V hardware by reaching milestones with its PolarFire system-on-a-chip (SoC) field-programmable gate array (FPGA) and Mi-V ecosystem.

AMD’s AI strategy comes into view with Xilinx, GPU, software plans

Source: The Register, June 13, 2022 [4]
AMD is now hoping to become a big player in the AI compute space. The chip designer believes it has the right silicon and software coming into place. AMD's vision is to provide a broad technology roadmap across training and inference that touches cloud, edge and endpoint.

AMD lines up alternate chips as it eyes a ‘post-exaflops’ future

Source: HPCwire, June 10, 2022 [5]
CEO Lisa Su says the company is diversifying risk by investing in a wider range of products. The company recently became the first chipmaker to pass the exaflops computing threshold. AMD's Frontier supercomputer topped the most recent Top500 supercomputing charts.

Co-constructing the Robot Prosperity Ecological Horizon Released the first domestic software and hardware integrated robot development platform

Source: Weixin Official Accounts Platform, June 14, 2022 [6]
Horizon established the AIoT & Universal Robot Division and launched the Horizon Hobot Platform. The platform provides an open source, soft-hard collaborative system solution, which is compatible with robots. Dr. Yu Kai said: "We hope to unite with a wide range of industry partners to jointly create a bright future for a prosperous robot ecology".

LEGO-like reconfigurable AI chip stacks sensing, processing layers

Source: Smart2zero, June 14, 2022 [7]
MIT engineers say they have developed a LEGO-like design for a stackable, reconfigurable artificial intelligence (AI) chip.

Partnership brings new capabilities to autonomous farming robot

Source: IoT World Today, June 14, 2022 [8]
Apex.AI is expanding its partnership with agricultural machinery manufacturer AGCO. The software development kit will bring autonomous capabilities to AGCO's Fendt Xaver concept vehicle. The battery-powered, cloud-controlled Xaver bots can autonomously plant seeds on farms 24 hours a day.

Improved interoperability between VPI and PyTorch

Source: NVIDIA Technical Blog, June 15, 2022 [9]
NVIDIA Vision Programming Interface (VPI) is a computer vision and image-processing software library from NVIDIA that enables you to implement algorithms that are accelerated on different hardware backends available on NVIDIA Jetson embedded devices and discrete GPUs

Papers

Reconfigurable heterogeneous integration using stackable chips with embedded artificial intelligence

Source: Nature, June 13, 2022 [10]
Three-dimensional heterogeneous integration with advanced packaging technologies could be used to improve data bandwidth among sensors, memory and processors. Such systems are limited by a lack of hardware reconfigurability and the use of von Neumann architectures. The technology could provide energy-efficient sensor computing systems for edge computing applications.

Analyze, debug, optimize: real-time tracing for perception and mapping systems in ROS 2

Source: arXiv, April 25, 2022 [11]
Perception and mapping systems are among the most computationally, memory, bandwidth intensive software components in robotics. We show how our novel opensource tracing tools and techniques for ROS 2 enable us to identify delays, bottlenecks and critical paths inside centralized, or distributed, perception and mapping system.

Research on heterogeneous acceleration platform based on FPGA

Source: ITM Web of Conferences, May 19, 2022 [12]
Academics have been struggling to cope with the increasing demands on computing power for massive data processing. Most of the transistors in CPUs are used to build cache memory and control units, which are not efficient for computing operations of massive data processing. The paper first discussed the advantages and technical requirements of FPGA acceleration based on the characteristics of the Spark cluster.

Previous Hardware Acceleration in Robotics Newsletters

Past ROS 2 Hardware Acceleration Working Group meetings

  1. Acceleration robotics launch ROBOTCORE™ to speed-up ROS 2 robots. (2022, June 16). businesswire. https://www.businesswire.com/news/home/20220615005050/en/Acceleration-Robotics-Launch-ROBOTCORE™-to-Speed-up-ROS-2-Robots ↩︎

  2. Crowe, S. (2022, June 10). ROS 2 driver now available for ABB's robot arms. The Robot Report. https://www.therobotreport.com/ros-2-driver-now-available-for-abbs-robot-arms/ ↩︎

  3. Hertz, J. (2022, June 11). Microchip Pushes First RISC-V-based SoC FPGA to Mass Production. allaboutcircuits. https://www.allaboutcircuits.com/news/microchip-pushes-first-risc-v-based-soc-fpga-to-mass-production/ ↩︎

  4. Martin, D. (2022, June 13). AMD’s AI strategy, from Xilinx and GPUs to software plans. The Register: Enterprise Technology News and Analysis. https://www.theregister.com/2022/06/13/amd_ai_strategy/ ↩︎

  5. Shah, A. (2022, June 10). AMD lines up alternate chips as it eyes a 'post-exaflops' future. HPCwire. https://www.hpcwire.com/2022/06/10/amd-lines-up-alternate-chips-as-it-eyes-a-post-exaflops-future/ ↩︎

  6. Horizon. (2022, June 14). 共建机器人繁荣生态 地平线发布国内首个软硬一体机器人开发平台. Weixin Official Accounts Platform. https://mp.weixin.qq.com/s/gX6nkzxdNgTVYK4ydYFolQ ↩︎

  7. Pell, R. (2022, June 14). LEGO-like reconfigurable AI chip stacks sensing, processing layers. Smart2zero. https://www.smart2zero.com/en/lego-like-reconfigurable-ai-chip-stacks-sensing-processing-layers/ ↩︎

  8. Evans, S. (2022, June 14). Partnership brings new capabilities to autonomous farming robot. IoT World Today. https://www.iotworldtoday.com/2022/06/14/partnership-brings-new-capabilities-to-autonomous-farming-robot/ ↩︎

  9. Hiremath, S., & Maximo, A. A. (2022, June 15). Improved interoperability between VPI and PyTorch. NVIDIA Technical Blog. https://developer.nvidia.com/blog/improved-interoperability-between-vpi-and-pytorch/ ↩︎

  10. Chanyeol, C., Hyunseok, K., Hoon, K. J., Kyu, S. M., Hanwool, T., Celesta, C. S., Min, S. J., Jiho, S., Kuangye, L., Yeongin, K., Eol, L. H., Doyoon, L., Jaeyong, L., Ikbeom, J., Subeen, P., Kanghyun, R., Hoon, B. S., Yifan, N., Hyun, K., … Jeehwan, K. (2022, June 13). Reconfigurable heterogeneous integration using stackable chips with embedded artificial intelligence. Nature. https://www.nature.com/articles/s41928-022-00778-y ↩︎

  11. Lajoie, P. Y., Bedard, C., & Beltrame, G. (2022, April 25). Analyze, Debug, Optimize: Real-Time Tracing for Perception and Mapping Systems in ROS 2. arXiv.org e-Print archive. https://arxiv.org/pdf/2204.11778.pdf ↩︎

  12. Yuang, M., & Jun, Y. (2022, May 19). Research on heterogeneous acceleration platform based on FPGA. ITM Web of Conferences. https://www.itm-conferences.org/articles/itmconf/pdf/2022/05/itmconf_cscns2022_01029.pdf ↩︎