| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
How is the performance of ndt matching in cuda cores?
Is it possible to use DLA OF xavier for yolo or ssd based detection?
Xavier has 512 cuda cores and DLA.
Do we have any requirement like minimum number of cuda cores required for real time ndt matching?
I assume you're talking about Autoware.ai because your question does not say Ai or Auto. The Xavier is not an officially-supported platform but the Autoware.ai stack is currently being built for arm64-based hardware (from source and in Docker containers). All you have to do to enable compilation with CUDA is set the environment variable AUTOWARE_COMPILE_WITH_CUDA=1. See these instructions for more information. However, as I said, it is not an officially-supported or tested environment so I wish you luck.
Asked: 2020-04-25 09:41:34 -0500
Seen: 633 times
Last updated: Apr 30 '20
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
