Nvidia boosts AI assistants on PCs with new GeForce RTX features.
Nvidia made a splash at Computex 2024, showcasing a suite of new RTX AI technologies designed to revolutionize AI assistants, digital humans, and content creation on laptops. Here’s a breakdown of the key announcements:
Project G-Assist: The AI Assistant of the Future
Project G-Assist is an RTX-powered AI assistant demo that offers context-aware help for PC games and applications. It leverages generative AI to understand players’ needs and provide assistance within the game itself. Here’s how it works:
| Feature | Description |
|---|---|
| Contextual Awareness | G-Assist analyzes voice or text inputs and game screen information to understand the situation. |
| Large Language Model (LLM) | A powerful LLM linked to a game knowledge database processes the information and generates tailored responses. |
| Personalized Support | G-Assist personalizes its responses based on the player’s current game session. |
Benefits of Project G-Assist:
- Improved Gameplay: Gain insights into creatures, items, lore, objectives, and boss strategies.
- Streamlined Workflows: Receive assistance with complex creative tasks within applications.
- Performance Optimization: G-Assist can configure your system for optimal performance and efficiency.
Nvidia ACE Comes to RTX AI PCs
Nvidia is bringing its digital human development platform, Nvidia ACE, to RTX AI laptops and workstations. This allows developers to create lifelike digital humans with capabilities like natural language understanding, speech synthesis, and facial animation.
Key benefits of Nvidia ACE on RTX AI PCs:
- Reduced Deployment Times: Leverage Nvidia NIM (inference microservices) to streamline development.
- High-Quality Inference: Enables local processing of complex AI tasks for smooth performance.
Collaboration with Microsoft on Windows Copilot
Microsoft and Nvidia are teaming up to bring new AI capabilities to Windows applications. This collaboration will provide developers with access to GPU-accelerated small language models (SLMs) for tasks like content summarization, generation, and automation.
Faster and Smaller AI Models with RTX AI Toolkit
The RTX AI Toolkit is a suite of tools and resources designed to empower developers to build and deploy AI models specifically for RTX AI PCs. Here’s what it offers:
- Model Customization: Open-source QLoRa tools allow for tailoring pre-trained models.
- Model Optimization: Nvidia TensorRT optimizes models for faster performance and reduced memory usage.
- Deployment Support: The Nvidia AI Inference Manager simplifies AI integration into PC applications.
Benefits of RTX AI Toolkit for Developers:
- Faster Performance: Achieve up to 4x faster performance compared to pre-trained models.
- Reduced Model Size: Models can be optimized to consume up to 3x less RAM.
- Simplified Integration: The AI Inference Manager streamlines the integration process.
Integration with Popular Creative Applications
Several software partners are integrating components of the RTX AI Toolkit into their applications. This will unlock new possibilities for AI-powered content creation.
Examples of RTX AI Toolkit Integration:
- Adobe Photoshop: Leverages TensorRT for faster performance and AI-powered capabilities.
- ComfyUI: RTX acceleration delivers a 60% performance improvement over the current version.
- Automatic1111: RTX acceleration streamlines workflows for Stable Diffusion users.
RTX Remix: A Boon for Modders
Nvidia is expanding the capabilities of RTX Remix, its modding platform for classic games. This update allows modders to create even more stunning remasters:
- Open-Source Toolkit: More of the platform is becoming open-source, empowering the modding community.
- Expanded Functionality: New features include streamlined asset replacement, scene relighting, and enhanced AI texture tools.
- REST API and SDK: Integration with other content creation tools and games is now possible.
Benefits of RTX Remix for Modders:
- Simplified Workflows: Streamlined asset replacement and scene relighting.
- Enhanced Capabilities: New AI texture tools and support for more file formats.
- Greater Integration: Live link with other tools and deployable renderer for broader use.
RTX Video Goes Beyond Browsers
Previously available only in web browsers, Nvidia RTX Video, the AI-powered video upscaling feature, is now available as an SDK for developers. This allows them to integrate AI for upscaling, sharpening, and HDR conversion within their applications.
Future of RTX Video:
- Video Editing Software: Upscale video to 4K and convert SDR to HDR in DaVinci Resolve and Filmora.
- VLC Media Player: Soon to offer RTX Video HDR capabilities for existing super-resolution features.