Drivers and Software updates for Nvidia’s recently unveiled GeForce RTX 5090 and RTX 5080 GPUs have arrived, introducing AI-driven enhancements designed to elevate audio and video quality for content creators and streamers.
Alongside the new hardware, a significant update to Nvidia’s Broadcast app brings fresh AI effects and improvements to existing tools. The primary focus is on improving audio fidelity using AI.
The launch of these GPUs comes after Nvidia’s previous unveiling at CES 2025, where new technologies like Reflex 2 were initially showcased.
Studio Voice and Virtual Key Light: New AI-Driven Tools
The Broadcast app update debuts two beta AI effects: “Studio Voice” and “Virtual Key Light.” Studio Voice aims to enhance microphone audio, with a goal to make low-quality, built-in mics sound comparable to professional devices.
A demonstration of “Studio Voice” from The Verge showcased a clear improvement in audio quality from a basic webcam microphone. The “Virtual Key Light” effect simulates studio lighting, altering a subject’s appearance as though illuminated by professional equipment.

While the test from The Verge suggests it might not be quite as impactful as “Studio Voice”, the objective is “to deliver even lighting, as if a physical key light was defining the form and dimension of an individual.” Due to the high processing demands both new features necessitate a GeForce RTX 4080 or higher GPU and are not recommended for gaming use.
Broadcast App Enhancements and Integration
Existing features within the Broadcast app have also received improvements. Background Noise Removal now provides enhanced voice quality, while “Eye Contact adds gaze stability and subtle random eye movements for a more natural appearance.” Furthermore, Virtual Background has enhanced foreground and background separation.
The user interface allows users to apply multiple effects simultaneously, featuring a side-by-side camera preview, a GPU utilization meter, and more. Developers can incorporate these effects directly into their applications through NVIDIA Maxine Windows software development kits (SDKs), or via an NVIDIA NIM microservice.
RTX 50 Series: Enhanced Hardware Performance
The RTX 5090 and RTX 5080 GPUs, built on the Blackwell architecture, feature fifth-generation Tensor Cores with FP4 support. This reduces the VRAM needed for generative AI models while also increasing overall performance, according to Nvidia.
The RTX 5090 has 32GB of GDDR7 memory, 1,792 GB/sec bandwidth, a 77% increase over the RTX 4090. Additionally, it includes three encoders and two decoders, reducing video export times by a third in comparison to the previous generation.
The RTX 5080 includes 16GB of GDDR7 memory and 960 GB/sec of bandwidth, a 34% increase over the RTX 4080. These hardware improvements translate to significant advantages for video editing and streaming, and as an example, Black Forest Labs’ FLUX models operate using less than 10GB of VRAM at FP4 precision, reduced from over 23GB with FP16.
Improved Workflow for Video and Streaming
The RTX 50 Series GPUs feature 4:2:2 hardware support and can decode up to 8K video at 75 fps or nine 4K sources at 30 fps, for smooth multi-camera editing. The RTX 5090 has three encoders and two decoders, allowing video exports to be 40% faster than the RTX 4090, and four times faster than the RTX 3090.
The ninth-generation NVENC provides a 5% video quality increase for HEVC and AV1 encoding, and the new AV1 Ultra Quality mode improves compression by 5% without sacrificing quality. Software such as Blackmagic Design’s DaVinci Resolve, and Wondershare Filmora already incorporate these enhancements.
Live streamers also benefit with the ninth-generation NVENC, with a 5% video quality improvement on platforms like Twitch, YouTube, and Discord, measured by comparing the bitrates that match the same video quality.
3D Rendering, Gaming Advancements, and Related News
The RTX 5090’s 32GB of memory provides substantial advantages for 3D artists, allowing them to work on large projects and across different platforms smoothly. Fourth-generation RT Cores allow 3D applications to run 40% faster.
DLSS 4 with Multi Frame Generation is now available in D5 Render and will be coming in February to Chaos Vantage. Furthermore, the Game Ready 572.16 driver also introduces Smooth Motion AI and upgrades to for a number of games and applications.
Moreover, Nvidia’s Reflex 2 technology with Frame Warp reduces latency by up to 75%. According to NVIDIA, “Every player action taken in a video game goes through a complex pipeline before being rendered on-screen, with each step introducing latency.“