Support Matrix#
For production, Holoscan for Media is designed to run on NVIDIA-Certified Systems. For full hardware, networking, and cluster platform requirements, refer to the System Requirements for Holoscan for Media.
Models#
Model name |
Model ID |
Publisher |
|---|---|---|
Active Speaker Detection |
active-speaker-detection |
NVIDIA |
Supported Video Resolutions#
Resolution |
Width × Height |
Notes |
|---|---|---|
4K UHD |
3840 × 2160 |
Supported |
1080p |
1920 × 1080 |
Default |
720p |
1280 × 720 |
Supported |
Set the resolution via video.width and video.height. The upstream sender and downstream receiver must be configured with the same resolution.
Optimized GPU Configurations#
GPU |
Precision |
|---|---|
L40S |
FP16 |
NVIDIA RTX PRO 6000 Blackwell Server Edition |
FP16 |
Pod Resource Requirements#
The following resources are required per Active Speaker Detection NIM media function pod:
Resource |
Requirement |
|---|---|
CPU |
12 cores |
Memory |
8 GiB |
Hugepages (2 Mi) |
8 GiB |
NVIDIA GPU |
1 |
Ensure the target node has sufficient resources and that the GPU and hugepages are configured before deploying. Node selector configuration is covered in each chart’s installation page.
Software#
NVIDIA Driver#
Prerequisite |
Version |
Reference |
|---|---|---|
NVIDIA Graphics Driver for Linux |
571.21+ |
Component Versions#
The Active Speaker Detection NIM on Holoscan for Media uses the following software components:
Component |
Version |
|---|---|
CUDA |
12.8.1 |
cuDNN |
9.7.1.26 |
TensorRT |
10.9.0.34 |
Triton Inference Server |
v2.56.0 |
DeepStream |
8.0 |
NVIDIA Media Gateway |
0.7.0 |