Isaac Sim Benchmarks#

This page contains key performance indicators (KPIs) for Isaac Sim, captured across different reference hardware and measured using the isaacsim.benchmark.services extension. It also contains a guide on how to collect the same KPIs on your hardware, to compare to our published performance specs.

GPU-Independent KPIs#

These KPIs measure Isaac Sim performance independent of the GPU on which Isaac Sim is running.

Note

These KPIs were measured on a standardized reference machine using an Intel i9-14900k CPU and 32GB of DDR5 RAM.

GPU-Independent KPIs#

Name

Definition

Units

Value

Binary package size (Windows)

Size of Windows binary package

GB

7.37

Binary package size (Ubuntu)

Size of Ubuntu binary package

GB

8.17

Docker container size

Size of Docker container before extraction on NGC

GB

pip package size

Size of pip package as downloaded

GB

Startup time (async)

Time from launching Isaac Sim executable to app ready appearing in logs

seconds

31.472 [1]
6.31 [2]

Startup time (non-async)

Time from initializing SimulationApp in standalone Python to app ready appearing in logs

seconds

263 [3]
4.43 [4]

GPU-Dependent KPIs#

These KPIs measure Isaac Sim performance on reference hardware, including frame rate for benchmark scenes and render rate for specific sensor combinations. KPIs are reported as the average KPI value across 600 frames.

Note

SDG KPIs are measured using two (2) 720p cameras in a blank scene with 500 objects generated each frame.

Note

ROS2 Render & Publishing Speed is measured using the reference Nova Carter asset, rendering all 4 stereo cameras, 1 3D lidar and 2 2D lidars, and publishing all possible messages.

Note

The following KPIs were measured with Motion BVH enabled. This is a rendering setting that allows for higher accuracy when modeling motion-related sensor effects and is enabled by default in Isaac Sim. This will cause slower rendering performance, and can be disabled for substantial performance gains if unneeded. Further details can be found in the Performance Optimization Handbook.

Note

These KPIs were measured on a standardized reference machine using an Intel i9-14900k CPU and 32GB of DDR5 RAM.

Hardware-Dependent KPIs#

Name

Definition

Units

Windows

Ubuntu

Full Warehouse Sample Scene Load Time

Wall-clock time to load Full Warehouse Sample Scene

Seconds

56.3

49.6

Full Warehouse Sample Scene FPS

Frame rate of Full Warehouse Sample Scene

Frames per second

78.56

74.85

Physics steps per second

Number of physics steps executed per wall-clock second with 10 O3dyn robots

Hz

52.91

59.28

Isaac ROS Sample Scene FPS

Frame rate of Isaac ROS Sample Scene

Frames per second

23.35

23.14

ROS2 render & publishing speed

Frame rate rendered and published via ROS2 bridge from Nova Carter ROS asset, per wall-clock second

Frames per Second

6.04

5.81

SDG images per second (simple)

Frame rate rendered by SDG per second, with only RGBD annotators enabled, per wall-clock second

Frames per second

6.07

8.49

SDG images per second (complex)

Frame rate rendered by SDG per second, with all annotators enabled, per wall-clock second

Frames per second

3.89

6.68

Note

These KPIs were measured on a standardized reference machine using an Intel i9-14900k CPU and 32GB of DDR5 RAM.

Hardware-Dependent KPIs#

Name

Definition

Units

Windows

Ubuntu

Full Warehouse Sample Scene Load Time

Wall-clock time to load Full Warehouse Sample Scene

Seconds

54.1

48.7

Full Warehouse Sample Scene FPS

Frame rate of Full Warehouse Sample Scene

Frames per second

95.51

82.58

Physics steps per second

Number of physics steps executed per wall-clock second with 10 O3dyn robots

Hz

42.11

44.76

Isaac ROS Sample Scene FPS

Frame rate of Isaac ROS Sample Scene

Frames per second

27.68

25.29

ROS2 render & publishing speed

Frame rate rendered and published via ROS2 bridge from Nova Carter ROS asset, per wall-clock second

Frames per second

7.43

7.91

SDG images per second (simple)

Frame rate rendered by SDG per second, with only RGBD annotators enabled, per wall-clock second

Frames per second

6.09

8.52

SDG images per second (complex)

Frame rate rendered by SDG per second, with all annotators enabled, per wall-clock second

Frames per second

4.28

6.62

Note

These KPIs were measured on a standardized reference machine using an Intel i9-14900k CPU and 32GB of DDR5 RAM.

Hardware-Dependent KPIs#

Name

Definition

Units

Windows

Ubuntu

Full Warehouse Sample Scene Load Time

Wall-clock time to load Full Warehouse Sample Scene

Seconds

55.9

46.2

Full Warehouse Sample Scene FPS

Frame rate of Full Warehouse Sample Scene

Frames per second

166.67

140.06

Physics steps per second

Number of physics steps executed per wall-clock second with 10 O3dyn robots

Hz

41.2

45.54

Isaac ROS Sample Scene FPS

Frame rate of Isaac ROS Sample Scene

Frames per second

47.87

45.41

ROS2 render & publishing speed

Frame rate rendered and published via ROS2 bridge from Nova Carter ROS asset, per wall-clock second

Frames per second

12.88

11.76

SDG images per second (simple)

Frame rate rendered by SDG per second, with only RGBD annotators enabled, per wall-clock second

Frames per second

6.08

8.52

SDG images per second (complex)

Frame rate rendered by SDG per second, with all annotators enabled, per wall-clock second

Frames per second

4.72

6.73

Note

These KPIs were measured on a standardized OVX machine using 2x Intel 8362 CPU and 1024GB of DDR4 RAM, on Ubuntu 22.04. Some KPIs are measured on multi-GPU configurations, typically for 2, 4, or 8 GPUs.

Hardware-Dependent KPIs by GPU Count#

Name

Definition

Units

x1

x2

x4

x8

Full Warehouse Sample Scene Load Time

Wall-clock time to load Full Warehouse Sample Scene

Seconds

89.4

117.2

90.4

92.9

Full Warehouse Sample Scene FPS

Frame rate of Full Warehouse Sample Scene

Frames per second

38.4

53.9

54.2

46.9

Physics steps per second

Number of physics steps executed per wall-clock second with 10 O3dyn robots

Hz

37.3

37.7

37.3

37.2

Isaac ROS Sample Scene FPS

Frame rate of Isaac ROS Sample Scene

Frames per second

13.6

25.4

31.9

32.9

ROS2 render & publishing speed

Frame rate rendered and published via ROS2 bridge from Nova Carter ROS asset, per wall-clock second

Frames per second

3.3

5.0

8.3

13.1

SDG images per second (simple)

Frame rate rendered by SDG per second, with only RGBD annotators enabled, per wall-clock second

Frames per second

3.2

2.7

2.7

2.6

SDG images per second (complex)

Frame rate rendered by SDG per second, with all annotators enabled, per wall-clock second

Frames per second

2.3

2.2

2.2

2.2

Note

These KPIs were measured on a standardized OVX machine using 2x Intel 8362 CPU and 1024GB of DDR4 RAM, on Ubuntu 24.04. Some KPIs are measured on multi-GPU configurations, typically for 2, 4, or 8 GPUs.

Hardware-Dependent KPIs by GPU Count#

Name

Definition

Units

x1

x2

x4

x8

Full Warehouse Sample Scene Load Time

Wall-clock time to load Full Warehouse Sample Scene

Seconds

99.1

99.7

90.0

92.3

Full Warehouse Sample Scene FPS

Frame rate of Full Warehouse Sample Scene

Frames per second

97.0

61.9

61.9

47.2

Physics steps per second

Number of physics steps executed per wall-clock second with 10 O3dyn robots

Hz

47.7

46.8

47.6

46.8

Isaac ROS Sample Scene FPS

Frame rate of Isaac ROS Sample Scene

Frames per second

30.1

43.3

38.5

34.5

ROS2 render & publishing speed

Frame rate rendered and published via ROS2 bridge from Nova Carter ROS asset, per wall-clock second

Frames per second

8.0

9.6

14.8

19.8

SDG images per second (simple)

Frame rate rendered by SDG per second, with only RGBD annotators enabled, per wall-clock second

Frames per second

3.0

2.6

2.6

2.6

SDG images per second (complex)

Frame rate rendered by SDG per second, with all annotators enabled, per wall-clock second

Frames per second

2.3

2.1

2.2

2.2

Measuring KPIs on Local Hardware#

Isaac Sim KPIs can be measured using the Python scripts provided in standalone_examples/benchmarks. The KPIs provided above are computed using the scripts as follows. Commands are providing in bash syntax (for Ubuntu); for Windows, replace .sh with .bat and \ for multiline commands to `.

  • Startup time (async): Measured as Runtime for phase: startup from the logs after running

    ./isaac-sim.sh --no-window --/app/quitAfter=200 --/app/file/ignoreUnsavedOnExit=1 \
  --enable isaacsim.benchmark.service

  • Startup time (non-async): Measured as Runtime for phase: startup from the logs after running

    ./python.sh standalone_examples/api/isaacsim.simulation_app/hello_world.py \
  --enable isaacsim.benchmark.services

  • Full Warehouse Sample Scene load time & FPS: Load time is measured as Runtime for phase: loading and FPS is measured as Mean FPS for phase: benchmark.

    ./python.sh standalone_examples/benchmarks/benchmark_scene_loading.py \
  --env-url /Isaac/Environments/Simple_Warehouse/full_warehouse.usd

  • Physics Steps per second: Physics steps per second is measured as 1000.0/Mean Physics Frametime for phase: benchmark. An additional physics argument is available to switch between CPU and GPU (default) physics

    ./python.sh standalone_examples/benchmarks/benchmark_robots_o3dyn.py \
  --num-robots 10 --num-gpus 1

  • Isaac ROS Sample Scene FPS: FPS is measured as Mean FPS for phase: benchmark.

    ./python.sh standalone_examples/benchmarks/benchmark_scene_loading.py \
  --env-url /Isaac/Samples/ROS2/Scenario/carter_warehouse_apriltags_worker.usd

  • ROS2 Render & Publishing Speed: FPS is measured as Mean FPS for phase: benchmark.

    ./python.sh standalone_examples/benchmarks/benchmark_robots_nova_carter_ros2.py \
  --num-robots 1 --enable-3d-lidar 1 --enable-2d-lidar 2 --enable-hawks 4

  • SDG images per second (simple): FPS is measured as Mean FPS for phase: benchmark.

    ./python.sh standalone_examples/benchmarks/benchmark_sdg.py \
  --num-cameras 2 --resolution 1280 720 --asset-count 100
  --annotators rgb distance_to_image_plane --skip-write

  • SDG images per second (complex): FPS is measured as Mean FPS for phase: benchmark.

    ./python.sh standalone_examples/benchmarks/benchmark_sdg.py \
  --num-cameras 2 --resolution 1280 720 --asset-count 100
  --annotators all --skip-write