NVIDIA RTX IO

A suite of technologies that enables rapid GPU-based loading and asset decompression.

Key Benefits

Faster Load Times

RTX IO delivers near instantaneous game loading. Object pop-in and stutter can be reduced, and high-quality textures can be streamed at incredible rates.

Lower CPU Utilization

It offloads many CPU cores’ worth of work to your NVIDIA® GeForce RTX™ GPU, delivering maximum overall system performance for next-generation games.

Smaller Assets

With lossless compression, game download and install sizes can be reduced, allowing gamers to store more games on their solid-state drive (SSD) while improving their performance.

Open Source

The GDeflate compression format will be open source, enabling developers to create ports and tools.

Higher Throughput with RTX IO

NVIDIA RTX IO on and off comparison — NVIDIA RTX™ IO lets compressed data be delivered to GPU memory with minimal staging in system memory. The GPU is utilized for decompression using GDeflate at high throughput. The CPU is freed to perform other auxiliary tasks.

End-to-End Throughput

RTX IO with GDeflate GPU decompression delivers IO-saturating performance on modern NVMe devices. When comparings end-to-end throughput for different decompressors through DirectStorage, using the DirectStorage GPU Decompression Benchmark, decompressing through ZLib on the CPU cannot saturate NVMe bandwidth, while GDeflate keeps NVMe saturated, with room to spare.

Comparison graph of end-to-dnd throughput for different decompressors

Works on All Platforms, Optimized for NVIDIA Ada

Every End User Benefits

RTX IO is designed to be open source. It works across various operating systems and graphics APIs, computer-accelerated APIs, and SDKs. All modern GPUs will benefit from the ability to speed up load times and reduce CPU utilization. Thanks to a bigger cache size and optimized streaming multiprocessor (SM) cores, the NVIDIA Ada Lovelace architecture is where RTX IO will deliver best-in-class performance.

RTX IO-Supported APIs

Microsoft DirectStorage

NVIDIA RTX IO plugs into Microsoft’s DirectStorage API, which is a next-generation storage architecture designed for gaming PCs equipped with state-of-the-art NVMe SSDs and the complex workloads that modern games require. Together, the streamlined and parallelized APIs, specifically tailored for games, dramatically reduce overhead and maximize performance and bandwidth from NVMe SSDs to RTX IO-enabled GPUs.

Vulkan

Two new NVIDIA Vulkan extensions to accelerate RTX IO are supported on both Linux and Windows. VK_NV_memory_decompression handles decompressing assets on GPU, and VK_NV_copy_memory_indirect handles copying decompressed data to Vulkan buffers/images. These extensions are released in the Vulkan 1.3.233 update and supported in the latest GRD release.
This enables Vulkan applications and games across multiple platforms with faster transfer of assets from disk to GPU memory. Standardizing on the use of GDeflate format in Vulkan allows cross compatibility across many games.