Performance

DeepStream application is benchmarked across various NVIDIA TLT and open source models. The measured performance represents end-to-end performance of the entire video analytic application considering video capture and decode, pre-processing, batching, inference, and post-processing to generate metadata. The output rendering has been turned off to achieve peak inference performance. For information on disabling the output rendering, see DeepStream Reference Application - deepstream-app chapter.

TLT Pre-trained models

Transfer Learning Toolkit (TLT) has a set of pretrained models listed in the table below. If the models below satisfy your requirement, you should start with one of them. These could be used for various applications in smart city or smart places. If your application is beyond the scope of these models, you can re-train one of the popular model architecture using TLT. The second table shows the expected performance of few of other TLT models. The table below shows the end-to-end performance on highly accurate pre-trained models from TLT. All models are available on NGC. These models are natively integrated with DeepStream and the instructions to run these models are in /opt/nvidia/deepstream/deepstream-5.0/samples/configs/tlt_pretrained_models/.

Performance - pretrained models

Jetson Nano

Jetson Xavier NX

Jetson AGX Xavier

T4

Model Arch

Inference resolution

Precision

GPU (FPS)

GPU (FPS)

DLA1 (FPS)

DLA2 (FPS)

GPU (FPS)

DLA1 (FPS)

DLA2 (FPS)

GPU (FPS)

PeopleNet- ResNet34

960x544

INT8

10

157

51

51

272

67

67

807

PeopleNet – ResNet18

960x544

INT8

14

218

72

72

384

94

94

1105

TrafficCamNet – ResNet18

960x544

INT8

19

261

105

105

464

140

140

1300

DashCamNet – ResNet18

960x544

INT8

18

252

102

102

442

133

133

1280

FaceDetectIR- ResNet18

384x240

INT8

95

1188

570

570

2006

750

750

2530

All the models in the table above can run solely on DLA. This saves valuable GPU resources to run more complex models. TLT also supports training on popular Detection and Segmentation architectures. To learn more about how to train with TLT, refer to the TLT documentation. These models are natively integrated with DeepStream. These are just reference models and are available to download from GitHub.

Performance - pretrained models- detection and segmentation

Jetson Nano

Jetson Xavier NX

Jetson AGX Xavier

T4

Model Arch

Inference resolution

Precision

GPU (FPS)

GPU (FPS)

DLA1 (FPS)

DLA2 (FPS)

GPU (FPS)

DLA1 (FPS)

DLA2 (FPS)

GPU (FPS)

YoloV3 – ResNet18

960x544

INT8

11

78

55

55

223

84

84

620

FasterRCNN – ResNet10

480x272

INT8

16

127

N/A

N/A

281

N/A

N/A

932

SSD – ResNet18

960x544

INT8

10.6

124

56

56

216

77

77

760

DSSD – ResNet18

960x544

INT8

9

66

45

45

189

67

67

586

RetinaNet – ResNet18

960x544

INT8

8.5

60

45

45

147

41

41

296

MaskRCNN – ResNet50

1344x832

INT8

0.6

5.4

3.2

3.2

9.2

4.5

4.5

24

Note

  • FasterRCNN model will not run efficiently on the DLA due to multiple layers not supported on the DLA.

  • All inferences on Jetson Nano is done using FP16 precision.

DeepStream reference model and tracker

DeepStream SDK ships with a reference DetectNet_v2-ResNet10 model and 3 ResNet18 classifier models. The detailed instructions to run these models with DeepStream are provided in the next section. The table below shows the performance of these models along with various trackers. DeepStream provides 3 reference trackers: IoU, KLT and NvDCF. For more information about trackers, See the Gst-nvtracker section in DeepStream Plugins Development Guide.

Performance - Deepstream reference models

Jetson Nano

Jetson Xavier NX

Jetson AGX Xavier

T4

Model Arch

Tracker

Inference resolution

GPU (FPS)

GPU (FPS)

DLA1 (FPS)

DLA2 (FPS)

GPU (FPS)

DLA1 (FPS)

DLA2 (FPS)

GPU (FPS)

DetectNetV2 – ResNet10

No Tracker

480x272

92

1079

510

510

1800

690

690

2520

DetectNetV2 – ResNet10

KLT

480x272

91

512*

277

277

892*

475

475

2316

DetectNetV2 – ResNet10

NvDCF

480x272

40

356*

170

170

812*

244

244

1293

DetectNetV2 – ResNet10 + ResNet18 classifiers

No Tracker

480x272 + 224x224

50

718

N/A

N/A

1386

N/A

N/A

2278

DetectNetV2 – ResNet10 (Python App)

No Tracker

480x272

92

1057

N/A

N/A

1744

N/A

N/A

2250

Note

  • * - Performance bottleneck identified which will be fixed in future release.

  • All inferences are done using INT8 precision except on Nano. On Nano, it is FP16.

  • Running inference simultaneously on multiple models is not supported on the DLA. You can only run one model at a time on the DLA.

To achieve the peak performance shown in the tables above, make sure the devices are properly cooled. For T4, make sure you use a server that meets the thermal and airflow requirements of T4. Apart from hardware setup, few options in the config file needs to be set to achieve the published performance. Take one of the config files from DeepStream SDK and make the required changes to replicate the peak performance below: Turn off output rendering, OSD, and tiler. OSD which stands for on-screen display is used to display bounding box, masks, and labels on the screen. If output rendering is disabled, creating bounding boxes is not required unless the output needs to be streamed over RTSP or saved to disk. Tiler is used to display the output in NxM tiled grid. Not needed if rendering is disabled. Output rendering, OSD and tiler use some % of compute resources and it can reduce the inference performance. To disable OSD, tiled display and output sink, make the following changes in the DeepStream config file.

  • To disable OSD, change enable to 0

    [osd]
enable=0

  • To disable tiling, change enable to 0

    [tiled-display]
enable=0

  • To turn-off output rendering, change the sink to fakesink.

    [sink0]
enable=1
#Type - 1=FakeSink 2=EglSink 3=File
type=1
sync=0

DeepStream reference model

Data center GPU - T4

This section describes configuration and settings for the DeepStream SDK on NVIDIA® Data center GPU - T4.

System Configuration

The system configuration for the DeepStream SDK is listed below:

T4 System configuration

System Configuration

Specification

CPU

Dual Intel® Xeon® CPU E5-2650 v4 @ 2.20GHz (48 threads total)

GPU

Tesla T4*

System Memory

128 GB DDR4, 2400MHz

Ubuntu

Ubuntu 18.04

GPU Driver

450.51

CUDA

10.2

TensorRT

7.0+

GPU clock frequency

1513 MHz

Application Configuration

Config file: source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.txt Change the following items in the config file:

  • The inference resolution of Primary GIE is specified in the samples/models/Primary_detector/resnet10.prototxt.

  • Change the ‘dim’ to 480x272.

  • Change batch size under streammux and primary-gie to match the number of streams.

  • Disable tiled display and rendering using instructions above.

  • Enable IoU tracker.

The application configuration for the DeepStream SDK is listed below:

T4 application configuration

Application Configuration

Specification

N×1080p 30 fps stream
sample_1080p_h265.mp4 (provided with the SDK) N=64
sample_1080p_h264.mp4 (provided with the SDK) N=39

Primary GIE

  • Resnet10 (480×272)

  • Batch Size = N

  • Interval=0

Tracker

Enabled. Processing at 480×272 resolution, IOU tracker enabled.

3 × Secondary GIEs
All batches size 32. Asynchronous mode enabled.

  • Secondary_VehicleTypes (224×224—Resnet18)

  • Secondary_CarColor (224×224—Resnet18)

  • Secondary_CarMake (224×224—Resnet18)

Tiled Display

Disabled

Rendering

Disabled

Achieved Performance The achieved performance of the DeepStream SDK under the specified system and application configuration are as follows:

Stream type

No. of Stream @ 30 FPS

CPU Utilization

GPU Utilization

H.265

64

8% to 10%

58%

H.264

39

5%

31%

Jetson

This section describes configuration and settings for the DeepStream SDK on NVIDIA Jetson™ platforms. JetPack 4.4 DP is used for software installation.

System Configuration

For the performance test:

  1. Max power mode is enabled: $ sudo nvpmodel -m 0

  2. The GPU clocks are stepped to maximum: $ sudo jetson_clocks

For information about supported power modes, see “Supported Modes and Power Efficiency” in the power management topics of NVIDIA Tegra Linux Driver Package Development Guide, e.g., “Power Management for Jetson AGX Xavier Devices.”

Jetson Nano

Config file: source8_1080p_dec_infer-resnet_tracker_tiled_display_fp16_nano.txt Change the following items in the config file:

  • Change batch size under ‘streammux’ and ‘primary-gie’ to match the number of streams.

  • Disable tiled display and rendering using instructions above.

  • Enable KLT tracker and change the tracker resolution to 480x272.

The following tables describe performance results for the NVIDIA Jetson Nano.
Jetson Nano application configuration

Application Configuration

Specification

N×1080p 30 fps streams
sample_1080p_h265.mp4 (provided with the SDK) N = 8
sample_1080p_h264.mp4 (provided with the SDK) N = 8

Primary GIE

  • Resnet10 (480×272) Asynchronous mode enabled

  • Batch Size = N

  • Interval = 4

Tracker

Enabled; processing at 480×272 resolution, KLT tracker enabled

OSD/tiled display

Disabled

Renderer

Disabled

Achieved Performance

Stream type

No. of Stream @ 30 FPS

CPU Utilization

GPU Utilization

H.265

8

39%

67%

H.264

8

39%

65%

Jetson AGX Xavier

Config file: source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.txt Change the following items in the config file:

  • The inference resolution of Primary GIE is specified in the ‘samples/models/Primary_detector/resnet10.prototxt’.

  • Change the ‘dim’ to 480x272.

  • Change batch size under streammux and primary-gie to match the number of streams.

  • Disable tiled display and rendering using instructions above.

  • Enable IOU tracker.

The following tables describe performance results for the NVIDIA Jetson AGX Xavier™.

Jetson Nano Pipeline Configuration (deepstream-app)

Application Configuration

Specification

N×1080p 30 fps streams
sample_1080p_h265.mp4 (provided with the SDK) N=45
sample_1080p_h264.mp4 (provided with the SDK) N=32

Primary GIE

  • Resnet10 (480×272) Asynchronous mode enabled

  • Batch Size = N

  • Interval = 0

Tracker

Enabled; processing at 480×272 resolution, IOU tracker enabled.

3× secondary GIEs

All batches are size 32.

  • Secondary_VehicleTypes (224×224—Resnet18)

  • Secondary_CarColor (224×224—Resnet18)

  • Secondary_CarMake (224×224—Resnet18)

OSD/tiled display

Disabled

Renderer

Disabled

Achieved Performance

Stream type

No. of Stream @ 30 FPS

CPU Utilization

GPU Utilization

H.265

45

22%

95%

H.264

32

19%

71%

Jetson NX

Config file: source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.txt Change the following items in the config file:

  • The inference resolution of Primary GIE is specified in the ‘samples/models/Primary_detector/resnet10.prototxt’.

  • Change the ‘dim’ to 480x272.

  • Change batch size under streammux and primary-gie to match the number of streams.

  • Disable tiled display and rendering using instructions above.

  • Enable IOU tracker.

The following tables describe performance results for the NVIDIA® Jetson NX™.

Jetson NX Pipeline Configuration (deepstream-app)

Application Configuration

Specification

N×1080p 30 fps streams
sample_1080p_h265.mp4 (provided with the SDK) N=23
sample_1080p_h264.mp4 (provided with the SDK) N=16

Primary GIE

  • Resnet10 (480×272) Asynchronous mode enabled

  • Batch Size = N

  • Interval = 0

Tracker

Enabled; processing at 480×272 resolution, IOU tracker enabled.

3× secondary GIEs

All batches are size 32.

  • Secondary_VehicleTypes (224×224—Resnet18)

  • Secondary_CarColor (224×224—Resnet18)

  • Secondary_CarMake (224×224—Resnet18)

OSD/tiled display

Disabled

Renderer

Disabled

Achieved Performance

Stream type

No. of Stream @ 30 FPS

CPU Utilization

GPU Utilization

H.265

23

55%

93%

H.264

16

45%

65%

Jetson TX2

Config file: source12_1080p_dec_infer-resnet_tracker_tiled_display_fp16_tx2.txt Change the following in the config file:

  • Change batch size under streammux and primary-gie to match the number of streams.

  • Disable tiled display and rendering using instructions above.

  • Enable KLT tracker and change the tracker resolution to 480x272.

The following tables describe performance results for the NVIDIA Jetson TX2.

Jetson TX2 Pipeline Configuration (deepstream-app)

Application Configuration

Specification

N×1080p 30 fps streams
sample_1080p_h265.mp4 (provided with the SDK) N=15
sample_1080p_h264.mp4 (provided with the SDK) N=14

Primary GIE

  • Resnet10 (480×272) Asynchronous mode enabled

  • Batch Size = N

  • Interval = 4

Tracker

Enabled; processing at 480×272 resolution, KLT tracker enabled

OSD/tiled display

Disabled

Renderer

Disabled

Achieved Performance

Stream type

No. of Stream @ 30 FPS

CPU Utilization

GPU Utilization

H.265

15

35%

47%

H.264

14

34%

43%

Jetson TX1

Config file: source8_1080p_dec_infer-resnet_tracker_tiled_display_fp16_tx1.txt Change the following in the config file:

  • Change batch size under streammux and primary-gie to match the number of streams.

  • Disable tiled display and rendering using instructions above.

  • Enable KLT tracker and change the tracker resolution to 480x272.

The following tables describe performance results for the NVIDIA Jetson TX1.

Jetson TX1 Pipeline Configuration (deepstream-app)

Application Configuration

Specification

N×1080p 30 fps streams
sample_1080p_h265.mp4 (provided with the SDK) N=13
sample_1080p_h264.mp4 (provided with the SDK) N=10

Primary GIE

  • Resnet10 (480×272) Asynchronous mode enabled

  • Batch Size = N

  • Interval = 4

Tracker

Enabled; processing at 480×272 resolution, KLT tracker enabled

OSD/tiled display

Disabled

Renderer

Disabled

Achieved Performance

Stream type

No. of Stream @ 30 FPS

CPU Utilization

GPU Utilization

H.265

13

56%

49%

H.264

10

43%

43%

Running applications using DLA

Jetson AGX Xavier and Jetson NX supports 2 DLA engines. DeepStream does support inferencing using GPU and DLAs in parallel. You can do this in separate processes or single process. You will need three separate sets of configs, configured to run on GPU, DLA0 and DLA1:

  • Separate processes:

When GPU and DLA are run in separate processes, set the environment variable CUDA_DEVICE_MAX_CONNECTIONS as 1 from the terminal where DLA config is running.

  • Single process:

DeepStream reference application supports multiple configs in the same process. To run DLA and GPU in same process, set environment variable CUDA_DEVICE_MAX_CONNECTIONS as 32: $ deepstream-app -c <gpuconfig> -c <dla0config> -c<dla1config>