Hi everyone,
I’m happy to call for the seventh Hardware Acceleration Working Group meeting. This time to be held on 2022-03-31T16:00:00Z (your browser’s time). In case you missed past meetings, here’s the ROS 2 Hardware Acceleration Working Group list with past recordings we are keeping. As for the working group resources, find them below:
ROS 2 Hardware Acceleration Working Group (HAWG) resources
-
Objective
: Drive creation, maintenance and testing of acceleration kernels on top of open standards (C++ and OpenCL) for optimized ROS 2 and Gazebo interactions over different compute substrates (including FPGAs, GPUs and ASICs). -
Reference hardware platform
: Kria K26 Adaptive SOM -
Meetings invite group
: ROS 2 Hardware Acceleration WG Google Group -
Instant messaging
: Matrix community (Matrix is an open network for secure, decentralized communication). -
Backlog management
: DEPRECATED (use official backlog): Phase 1: tools, examples, benchmarking and first demonstrators · GitHub -
Github organization
: ros-acceleration -
Meeting minutes
: HAWG minutes -
Discourse tag
: wg-acceleration -
Previous meetings
:
The meeting will start reviewing the progress of the WG. In particular, after last month’s discussion around a “Methodology for ROS 2 Hardware Acceleration”, we’ll touch on a) updates of REP-2008 to reflect the methodology for ROS 2 Hardware acceleration (#22), b) perception hardware acceleration across silicon architectures, wherein we’ll compare results of hardware accelerating the simple perception graph of #22 across FPGA and GPU commercial solutions while discussing potential improvements, and c) ROSCon 2022 Hardware Acceleration Workshop (call for workshops just announced) !
The session will also include a guest talk provided by Hyunjong Choi, Postdoctoral Researcher at University of California Riverside, which together with his team led PiCAS: New Design of Priority-Driven Chain-Aware Scheduling for ROS2. Hyunjong Choi’s research goal is to design safe, predictable, and efficient computing systems, especially in the presence of dynamic timing constraints. He’ll tell us about how although ROS2 claims to enhance the real-time capability, ensuring predictable end-to-end chain latency still remains a challenging problem he explored. To address it, their group propose a new priority-driven chain-aware scheduler for the ROS 2 wherein callbacks are prioritized based on the given timing requirements of the corresponding chains so that the end-to-end latency of critical chains can be improved with a predictable bound. They demonstrate interest results with NVIDIA Xavier NX. Worth listening!
The meeting is initially scheduled to be 60 minutes. Coordinates for the meeting below (we’re switching to Google Meet, which simplifies my life organizing things):
- Time: 2022-03-31T16:00:00Z
- Coordinates: https://meet.google.com/hbh-pfvj-nmw
- Preliminary agenda:
- Welcome and newcomers intro
- Working group progress review:
- #22 Update REP-2008 in alignment with Methodology
- Perception hardware acceleration across silicon architectures
- Robotic Processing Unit (RPU)
- ROSCon 2022 Hardware Acceleration Workshop
- (other topics)
- Guest talk: Hyunjong Choi, PiCAS: New Design of Priority-Driven Chain-Aware Scheduling for ROS2
- Q&A