Generative AI is revolutionizing Windows applications and games by enabling new capabilities, such as unscripted, dynamic NPCs, novel works of art, and significant frame rate boosts in gaming. As generative AI’s applications and capabilities expand, the demand for computing power to support these workloads increases. Hence comes into play Hybrid AI, which is an amalgamation of onboard AI acceleration of NVIDIA RTX and cloud-based GPUs to meet these requisites effectively.
Hybrid AI: A Harmonious Solution
Hybrid AI synergizes local PC and workstation computing with cloud scalability, offering flexibility to optimize AI workloads based on specific use cases, cost, and performance:
- Local AI on RTX GPUs:
- Delivers high performance and low latency.
- Always available, even offline.
- Cloud-based AI:
- Runs larger models and scales across many GPUs.
- Serves multiple clients simultaneously.
By leveraging both local and cloud resources, Hybrid AI ensures AI tasks run optimally, whether locally or in the cloud, and accelerates these tasks using NVIDIA GPUs and AI stacks, including TensorRT and TensorRT-LLM. This results in reduced wait times and enhanced AI-powered features for users.
Tools and Technologies Supporting Hybrid AI
Several tools and technologies from NVIDIA support hybrid AI workflows for creators, gamers, and developers:
For Creators: Dream in the Cloud, Realize Locally
Generative AI empowers artists to ideate, prototype, and brainstorm new creations. For example:
Generative AI by iStock, powered by NVIDIA Edify:
- Provides a generative photography service built for and with artists.
- Offers extensive tools to explore styles, and variations, modify parts of an image, or expand the canvas.
- Allows artists to ideate multiple times and bring concepts to life quickly.
Once the creative concept is ready, artists can utilize RTX-powered PCs and workstations for AI acceleration in over 125 top creative apps, including Photoshop, DaVinci Resolve, and Blender.
For Gamers: Hybrid ACE Revolutionises NPC Interaction
Hybrid AI, through NVIDIA ACE, enables interactive PC gaming by integrating generative AI models into digital avatars on RTX AI PCs. This allows developers to create NPCs that can understand and respond to human player text and speech in real time, enhancing the gaming experience.
For Developers: A Hybrid Tool That Runs Anywhere
Hybrid AI assists developers in building and tuning new AI models. NVIDIA AI Workbench helps developers:
- Create, test, and customize pre-trained generative AI models and LLMs on RTX GPUs.
- Access repositories including Hugging Face, GitHub, and NVIDIA NGC.
- Reproduce, collaborate on, and migrate projects easily.
- Scale up projects when additional performance is needed, and bring them back to local RTX systems for inference and light customization.
The Hybrid RAG Workbench project allows developers to:
- Create customizable RAG applications.
- Embed documents locally and run inference on local RTX systems or cloud endpoints.
- Adapt projects to use various models, endpoints, and containers.
- Quantize models to run on their preferred GPU.
Conclusion
Hybrid AI powered by NVIDIA GPUs offers a seamless integration of local and cloud resources, providing creators, gamers, and developers with optimized AI workflows. This enables the realization of cutting-edge AI-powered features, driving innovation and performance across a wide range of applications. Whether developing new AI models, creating art, or enhancing gaming experiences, hybrid AI solutions provide the best of both worlds for AI enthusiasts and professionals.
