DirectX Raytracing 1.2 announced with huge visual and performance enhancements

Ray tracing was first announced for gaming in March 2018 for the DirectX 12 API at the Game Developers Conference (GDC). However, the first game using this ray tracing technology, Battlefield V, wasn't released until November 2018. The reflections in the title were very impressive, and fans loved it, but this technology would only work when used with Nvidia's RTX 20 series GPUs at the time.

Since then, almost every modern graphics card from AMD, Nvidia, or Intel supports ray tracing. However, turning on this rendering technique significantly reduces a game's performance. And that hasn't changed much, even in 2025. Thus, Microsoft has been working on improving it for a while.

At GDC 2025, the Redmond giant finally announced DirectX Raytracing 1.2. The latest iteration of DirectX ray tracing promises to increase performance dramatically while delivering enhanced graphics and visuals.

In this article, we analyze the new DirectX Raytracing 1.2 and explain everything there is to know about it.

Note: Some aspects of this article are subjective and reflect the writer's opinions.

---

Massive performance improvements in ray tracing

DirectX Raytracing 1.2 preview will release soon (Image via Microsoft)

Microsoft announced two new revolutionary techniques with the release of DirectX Raytracing 1.2:

• Opacity micromaps: This technique optimizes how rays interact with the geometry of a game world by micromanaging the opacity data and shader invocations. It can deliver as much as 2.3 times better performance and lower rendering costs without affecting the visual quality.
• Shader execution reordering: The DirectX technique logically clumps similar shader executions in groups to reduce the burden on the rendering pipeline and decrease divergence. It can lead to as much as 2x performance enhancement in ray tracing shader processing.

These two techniques promise to improve ray tracing performance drastically and make it possible to run full path-traced games on more graphics cards.

---

Next-gen photorealistic graphics with neural rendering & cooperative vectors in DirectX

RTX Neural Shaders (Image via Nvidia)

Neural rendering seems to be the future of gaming graphics, and Nvidia is at the forefront of this technology. Neural textures allow developers to deliver photorealistic visuals in games that not only look much better but also have a smaller memory footprint.

It could be a game-changer for graphics cards with lower memory – like the Nvidia RTX 4060, which has only 8GB and struggles with some games.

The use of neural textures was first announced by Nvidia at CES 2025, followed by AMD a month later. However, Microsoft will now integrate it into the Shader Model 6.9, which will be released alongside DirectX Raytracing 1.2. This will ensure every GPU vendor can use this technology without restrictions.

Here are some key benefits of using neural rendering and cooperative vectors:

• Neural Texture Block Compression: This compression and decompression technique dramatically reduces texture size without any noticeable reduction in visual quality. When combined with cooperative vectors, it delivers a 10-time performance improvement in data inferencing.
• Real-time path tracing with neural supersampling and de-noising: DirectX Raytracing 1.2 makes real-time path tracing accessible to more graphics cards by reducing its performance cost. Neural supersampling and de-noising techniques will further elevate the graphics and performance.
• Neural Shading SDK will support DirectX and utilize cooperative vectors: Nvidia has already announced its Neural Shading SDK, which will support DirectX and cooperative vectors to provide greater visual realism without increasing rendering costs.

Microsoft has also updated its DirectX profiler debugger, PIX, to better support the new DirectX Raytracing 1.2 features. Furthermore, its user interface received updates to provide a better experience to developers.