NVIDIA’s Holoscan Allows Developers to Integrate AI Into Live Media

NVIDIA recently introduced Holoscan for Media, a new platform designed to streamline AI integration into live media applications. This software-defined system, which has just begun its limited release, enables media companies to operate their live media pipelines and AI processes on a single infrastructure. The platform is engineered to simplify the process for developers incorporating AI capabilities into their real-time media projects.

AI revolutionizing broadcast industry: NVIDIA’s Holoscan for Media platform is transforming content creation, distribution, and consumption by simplifying AI integration into live media applications.

Key features of Holoscan for Media:

• Launched in limited availability, it allows developers to easily integrate AI into live media applications
• Enables media companies to run live media pipelines on the same infrastructure as AI
• Provides a software-defined, AI-enabled platform for live media production
• Seamlessly integrates with NVIDIA’s suite of AI software development kits (SDKs)

Addressing industry challenges:

• Simplifies the complex process of building robust AI software stacks for live media application development
• Reduces barriers to deployment by eliminating the need for extensive expertise and resources
• Overcomes limitations of traditional software development tied to dedicated hardware

Deployment flexibility and efficiency:

• Cloud-native architecture allows applications to run from anywhere
• Eliminates the need for separate builds for different deployment types (on-premises, cloud, or edge)
• Includes Precision Time Protocol for audio-video synchronization and Networked Media Open Specifications for seamless communication between applications

RTX AI PCs and workstations:

• Complement Holoscan for Media by offering a robust foundation for developing immersive media experiences
• Provide access to the CUDA ecosystem and NVIDIA SDKs optimized for media and AI workloads
• Enable seamless transition of applications from workstation to deployment environments

NVIDIA AI Enterprise:

• Offers a comprehensive suite of AI software, tools, and frameworks optimized for NVIDIA GPUs
• Includes popular frameworks like TensorFlow, PyTorch, and RAPIDS for streamlined deployment
• Enables developers to build advanced AI capabilities directly into media applications

Showcasing AI-powered video editing tools:

• Blackmagic Design’s DaVinci Resolve 19 Studio introduces new AI features:
  • IntelliTrack AI for stabilizing footage and automatic audio panning
  • UltraNR for AI-accelerated denoising
  • RTX Video Super Resolution for sharpening low-resolution video
  • RTX Video HDR for enhancing standard dynamic range video to HDR10 color spaces

Topaz Video AI integration with Adobe After Effects:

• Introduces a plug-in for video upscaling and frame interpolation
• Allows users to access AI upscaling tools directly in After Effects compositions
• Features dramatic slow-motion video creation with the Apollo AI model
• Excels at upscaling low-resolution assets for larger projects

Performance enhancements:

• DaVinci Resolve AI effects are accelerated on RTX GPUs by TensorRT
• Topaz Video AI plug-in is accelerated on RTX GPUs, boosting AI performance by up to 70%
• Future updates promise further TensorRT performance improvements and efficiency optimizations

Implications for the broadcast industry: The integration of AI technologies like NVIDIA’s Holoscan for Media and RTX-powered tools is set to streamline workflows, enhance content quality, and unlock new creative possibilities for broadcasters and content creators.