Tech Forums > 3D-Tech News Around The Web

NVIDIA OptiX 2.6 /3.0 beta Available

(1/1)

Stefan:

--- Quote ---OptiX 2.6 is now available, bringing support for Kepler generation GPUs and LLVM optimizations for compatibility with CUDA 4.2 and greater. This version has been thoroughly tested and should be your choice for stable development,

    Kepler GPU SUpport
    OptiX now supports all sm_30 Kepler devices (GK104 and GK107).

    CUDA LLVM Optimizations
    CUDA 4.1 and 4.2 utilize a newer LLVM-based compiler, which generates PTX code with different performance characteristics than prior versions.



Version 3 of OptiX has developed in parrallel with version 2.6, and is available for developers wanting to examine the latest OptiX features. This important release include inlucdes support for:

    Callable Programs
    Enables the assignment of arbitrary CUDA functions to OptiX variables, enabling shade trees, referenced procedurals, and a far more diverse set of shading and rendering algorithms.

    OptiX-CUDA Interop
    Allows efficient sharing of data between OptiX and other CUDA applications.

    Bindless Textures
    Textures can now have indirect access by using CUDA RT_TEXTURE and storing this handle within OptiX variables and buffers.

    TCC Driver Support
    OptiX is now compatible with NVIDIA Tesla boards using the TCC driver for either single or multiple GPU usage, allowing OptiX applications to be fully compatible with NVIDIA Maximus system configurations.

    Improved Performance
    When using GL interop input buffers with multiple GPUs.
    Far Faster SBVH Builds
    Performance for SBVH has improved by approximately 800%
    Better Multi-GPU Load Balancing
    The algorithm for partitioning work across multiple GPUs has improved, increasing multi-GPU scalability.
    More Efficient Compilation
    The need to compile due to scene changes is far less often than in previous releases.
    New Texture Wrap Modes
    For RT_WRAP_MIRROR and RT_WRAP_CLAMP_TO_BORDER

    New SDK Samples
    Demonstrating: Callable Programs, Multiple Importance Sampling, Ray Differentials and CUDA Interop


--- End quote ---

Navigation

[0] Message Index