Attention

As of June 30, 2025, the Isaac ROS Buildfarm for Isaac ROS 2.1 on Ubuntu 20.04 Focal is no longer supported.

Due to an isolated infrastructure event, all ROS 2 Humble Debian packages that were previously built for Ubuntu 20.04 are no longer available in the Isaac Apt Repository. All artifacts for Isaac ROS 3.0 and later are built and maintained with a more robust pipeline.

Users are encouraged to migrate to the latest version of Isaac ROS. The source code for Isaac ROS 2.1 continues to be available on the release-2.1 branches of the Isaac ROS GitHub repositories.

The original documentation for Isaac ROS 2.1 is preserved below.

Performance Summary

The performance results of benchmarking the Isaac ROS packages on both Jetson and x86_64 platforms are below:

Node
Input Size
AGX Orin
AGX Xavier
Orin NX
Orin Nano 8GB
x86_64 w/ RTX 4060 Ti
720p

40 ms
58 ms
74 ms
65 ms
37 ms
720p

9.7 ms
14 ms
14 ms
21 ms
6.4 ms
576p

1.2 ms
3.8 ms
1.6 ms
2.3 ms
0.32 ms
576p

43 ms
120 ms
98 ms

28 ms
576p

4.1 ms
15 ms
6.2 ms
7.4 ms
4.4 ms
288p

5.6 ms
7.3 ms
5.6 ms
6.9 ms
4.2 ms
1080p

11 ms
18 ms
19 ms
26 ms
7.2 ms
1080p

5.4 ms
5.9 ms
5.3 ms
6.4 ms
3.9 ms
VGA

22 ms
80 ms
56 ms
81 ms
11 ms
VGA

23 ms
830 ms
540 ms
790 ms
10 ms
PeopleSemSegNet
544p

3.4 ms
12 ms
5.6 ms
6.4 ms
2.2 ms
PeopleSemSegNet
544p

4.0 ms
12 ms
5.7 ms

2.0 ms
VGA

1.6 ms
3.6 ms
2.3 ms

0.66 ms
~50 sq. m

64 ms
290 ms
140 ms
200 ms
20 ms
1080p

9.5 ms
12 ms


3.1 ms
I-frame Support
1080p

12 ms
19 ms


4.8 ms
P-frame Support
1080p

11 ms
13 ms


4.1 ms
Graph
Input Size
AGX Orin
AGX Xavier
Orin NX
Orin Nano 8GB
x86_64 w/ RTX 4060 Ti
720p

14 ms
21 ms
20 ms
26 ms
9.5 ms
576p

41 ms
130 ms
95 ms
110 ms
30 ms
VGA

5.7 ms
16 ms
7.4 ms
12 ms
14 ms
VGA

33 ms
160 ms
120 ms

15 ms
576p

20 ms
93 ms
42 ms

11 ms
288p

13 ms
20 ms
16 ms

12 ms
1080p

15 ms
21 ms
21 ms
31 ms
3.0 ms
544p

11 ms
19 ms
15 ms
22 ms
5.6 ms
PeopleSemSegNet
544p

8.1 ms
12 ms
9.6 ms
13 ms
5.5 ms

Methodology

The performance numbers provided above are measured using Isaac ROS Benchmark. Configuration files for reproducing results are provided in the launch scripts for each benchmark here.

Frames per second (fps) are measured using the maximum sustained framerate for the accelerated computational pipeline on each listed platforms. Performance measured includes input node → graph of node(s) under performance test → output node where the publishing rate of the input node is auto-tuned to discover the peak throughput dropping <5% of the frames. Average fps computed over 5 runs, discarding minimum and maximum runs; latency measured at 30hz publishing rate matching sensor capture rate.