Build AI-Powered Games with NVIDIA DLSS 4.5, RTX, and Unreal Engine 5

2026-04-30 · Source: NVIDIA Technical Blog · Field: Technology & Digital — Gaming & Interactive Media, Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, medium

Summary

NVIDIA has released DLSS 4.5 with Dynamic Multi Frame Generation, Multi Frame Generation 6X, and a second-generation transformer model for Super Resolution, now available to game developers via SDK. This update, building on DLSS 4's adoption in over 250 games, extends DLSS technologies to more than 700 games and applications. Additionally, NVIDIA introduced a TensorRT for RTX plugin for Unreal Engine's Neural Network Engine (NNE), offering 1.5x performance improvements over DirectML for AI workloads. The company also unveiled Kimodo, a research project for generating realistic 3D character animation from text or keyframes, and a guide for using ComfyUI to create pre-production assets on RTX GPUs. New GDC and GTC sessions on RTX neural rendering and AI, plus a webinar on path-traced hair in Unreal Engine 5.7, are also highlighted.

Key takeaway

For game developers aiming to enhance visual fidelity and streamline production, integrating NVIDIA's new DLSS 4.5 SDK can significantly boost frame rates and image quality. Additionally, adopting the TensorRT for RTX plugin for Unreal Engine's NNE will accelerate AI workloads by 1.5x, while exploring Kimodo and ComfyUI offers pathways to more efficient motion generation and pre-production asset creation. Your team should evaluate these tools to maintain a competitive edge in real-time rendering and animation.

Key insights

NVIDIA's latest tools enhance game development with advanced AI rendering, motion generation, and asset creation.

Principles

• AI-driven rendering improves image quality and frame rates.
• Efficient AI deployment accelerates in-engine workloads.
• Generative AI streamlines animation and asset pipelines.

Method

DLSS 4.5 integrates a second-generation transformer model for Super Resolution and Dynamic Multi Frame Generation. TensorRT for RTX accelerates Unreal Engine NNE inference. Kimodo synthesizes 3D motion from text or keyframes.

In practice

• Integrate DLSS 4.5 SDK for higher frame rates.
• Use TensorRT for RTX plugin for 1.5x AI performance.
• Explore Kimodo for scalable animation workflows.

Topics

• NVIDIA DLSS 4.5
• Unreal Engine 5
• TensorRT for RTX
• NVIDIA Kimodo
• Generative AI

Code references

• NVIDIA/NVIDIA-GenAI-Creator-Toolkit

Best for: Machine Learning Engineer, Computer Vision Engineer, Software Engineer, AI Engineer

Related on AIssential

• What’s New for Game Developers in NVIDIA RTX: DLSS 4.5 for UE5 and Multilingual AI Characters — NVIDIA's RTX ecosystem updates enable more immersive games via multilingual AI characters and advanced rendering techniques.
• NVIDIA and ComfyUI Streamline Local AI Video Generation for Game Developers and Creators at GDC — NVIDIA's GDC updates enhance local AI video generation with simplified interfaces and significant performance gains on RTX GPUs.
• NVIDIA RTX Innovations Are Powering the Next Era of Game Development — NVIDIA's GDC 2026 announcements advance real-time ray tracing, on-device AI, and cloud-based game development infrastructure.
• Enhancing Real-Time Development Workflows with High GPU Memory — Modern real-time development and generative AI workflows require immense VRAM, often exceeding 30-35 GB for complex scenes.
• Nvidia CEO tries to explain why DLSS 5 isn’t just “AI slop” — Nvidia's DLSS 5 is an artist-guided generative AI tool for game enhancement, not post-processing, reflecting a broader strategy of co-design and ecosystem influence.
• NVIDIA Levels Up Local AI Agents Across RTX PCs and DGX Spark — NVIDIA is driving secure, high-performance local AI agent adoption across its RTX and DGX hardware and software platforms.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by NVIDIA Technical Blog.