This is a repost of a question which was in the wrong location:
Do you have any collection of information about CUDA with recipe for various computing solutions?
Iām new to CUDA. Open to ideas of use cases for the followings:
- Real-time vision detection (hobby project)
- Refining and executing models for multi-lingual transcription of videos (research work)
- Video transcoding pipeline to regenerate video for distributions in different format/resolutions (work)
thanks
@kcsham Here is where we can continue this discussion! Could you please give a little bit more information regarding what you mean regarding ācomputing solutionsā?
Currently, one thing that comes to mind is our Hopper H100 Confidential Computing Solution (NVIDIA Confidential Computing - NVIDIA Docs). This information goes over how you can ensure confidentiality (encryption and authentication) of your data from prying eyes (physically or remotely)
We also have our High Performance Computing (HPC) for things like high-complexity simulations, etc.
Happy to keep talking based on what youāre looking for!
1 Like
You can take a look at our computer vision section: Explore Computer Vision Software & Cloud Platform | NVIDIA . Personally for a hobby/getting started I recommend looking at DeepStream SDK; it is generally best for āgetting it done fastā as it has many pre-built plugins for you to use.
For Audio transcription, you could take a look at Audio Transcription from NVIDIA LaunchPad This is a free lab where you can try our NVIDIA Riva speech recognition!
For video transcoding, we have FFmpeg plugins, which can be found, built, and integrated into this open-source, industry-standard video tool. Please see this for instructions (Using FFmpeg with NVIDIA GPU Hardware Acceleration - NVIDIA Docs). This plugin you might notice is part of our larger Video Codec SDK, which is our main area for doing transcoding/modification. The main document page for that can be found here: NVIDIA Video Codec SDK v12.1 (Latest Release) - NVIDIA Docs
1 Like