RTX Memory Utility (RTXMU) Available Now
Reducing Memory Consumption with an Open Source Solution
Real-time ray tracing has advanced the art of lighting in video games, but it’s a computationally expensive process. Aiming to reduce these costs, NVIDIA has developed a memory utility that combines both compaction and suballocation techniques to optimize and reduce memory consumption of acceleration structures. We’ve turned this solution into an SDK called RTXMU, and we are making it available as an open source release today. It’s built to support any DXR or Vulkan Ray Tracing application.
Compaction of acceleration structures with RTXMU eliminates any wasted memory from the initial build operation. For applications using RTXMU, NVIDIA RTX cards get a ~50% reduction in memory footprint. Additionally, suballocating acceleration structure buffers with RTXMU prevents fragmentation and wasted space. Scenes with thousands of small unique BLAS benefit greatly from suballocation.
How Can RTXMU Help You, Right Away?
RTXMU is easy to integrate, and it provides benefits immediately.
A suballocation and compaction memory manager takes significant engineering time to validate. RTXMU reduces the time it takes for a developer to integrate compaction and suballocation into an RTX title.
RTXMU also abstracts away the memory and compaction state management of the BLAS, and manages all barriers required for compaction size readback and compaction copies.
Diving a bit deeper, RTXMU uses handle indirection to BLAS data structures to prevent any mismanagement of CPU memory, which could include accessing a BLAS that has already been deallocated or doesn’t exist. Also, suballocation gives the benefit of less TLB (Translation Lookaside Buffer) misses by packing more BLASes into 64 KB or 4 MB pages.
Put simply, RTXMU will make your real-time ray traced games and applications run better, without significant effort on your part.
Where can I get RTXMU?
RTXMU is an open source SDK available today, and an update will come this week. For tips on deployment, check out our RTXMU getting started blog.
You can learn more about NVIDIA RTXMU at developer.nvidia.com/RTXMU.