Your Browser is Out of Date

Nytro.ai uses technology that works best in other browsers.
For a full experience use one of the browsers below
Developing and Deploying Vision AI with Dell and NVIDIA Metropolis on A30 GPU
Executive summary Object detection throughput using a R7515 Dell PowerEdge server with an NVIDIA A30 GPU Dell PowerEdge R7515 Server Specifications Conclusion
Conclusion
References

Conclusion

Conclusion

  • High accuracy video stream object detection and throughput is achieved using pruned 8-bit integer models exported by NVIDIA’s TAO toolkit running on a Dell PowerEdge R7515 server with an A30 GPU. Using NVIDIA’s DeepStream and Triton software to orchestrate multiple streams of video on the Dell R7515 produces more than 800 frames per second of object detection on the KITTI data set. A total of more than 2200 frames of video throughput in the streams being processed can be supported by performing inference on every fourth frame.