Questions tagged [cuda]
The cuda tag has no summary.
22 questions
0 votes
0 answers
78 views
What does `cudaFilterModeLinear` do for the 2D texture?
I am trying to reduce the number of loads and computation for the vertex normal LERP with barycentric coordinates. As I understand it, when implementing it in CUDA, CUDA texture can be used, since <...
- 1,462
2 votes
1 answer
437 views
How is Shader Execution Reordering implemented under the hood?
I am writing a software path tracer from scratch in CUDA (for learning purposes, without resorting to any higher-level graphics API like OptiX), and it is well-known that path tracing can have ...
- 1,462
1 vote
1 answer
180 views
Why is my ray tracer not accelerated by CUDA?
Summary: I ported a simple path tracer smallpt into CUDA but the performance doesn't boost. reproduce results: build and run the CPU version: ...
- 203
0 votes
0 answers
120 views
What performance to expect from an optimised pathtracer?
I am currently writing a fairly basic pathtracer as a compute shader in C++/CUDA, which has been running on a GTX 1080, and I have noticed that when logging some statistics, I get around ~1 billion ...
0 votes
1 answer
134 views
What is in a GPU driver?
There are 3 kinds of code that are contained in a GPU-utilizing program: CPU code. CPU-to-GPU or GPU-to-CPU bridge code. This is what DirectX actually is. GPU code. In NVidia speak, this is called a &...
1 vote
1 answer
1k views
On modern GPUs, what hardware do vertex/pixel shaders have access to that compute shaders don't?
Suppose that someone implemented rendering for a video game using entirely GPGPU apis, like compute shaders or CUDA. Would it be possible to achieve the same performance as standard graphics pipelines ...
- 113
3 votes
1 answer
7k views
Performance difference in OpenGL compute shader vs Vulkan Compute Shader vs Cuda/OpenCL and others
I have been told there is no performance difference if you are skilled enough when it comes to gpu computing. And cuda only performs better because nvidia wants it to. I have also read graphic api's ...
- 33
1 vote
1 answer
239 views
Why is the image from the CUDA version not as sharp as the CPU version?
I am extending Roger Allen's Ray Tracing in One Weekend in CUDA (and also here) by following Peter Shirley's Ray Tracing: The Next Week; specifically Image Texture Mapping. The image I obtain from ...
- 319
3 votes
1 answer
233 views
Why does my image has a square pattern in cuda path tracer?
I am writting a path tracer in cuda following the Shirley's book. I am having a square pattern as an output. The size of the block change if I change the block dimensions. For this image, the block ...
- 446
2 votes
0 answers
2k views
Groupshared memory and parallel reduction over multiple kernel dispatch
I am doing some parallel reduction and mostly following these nVidia slides. However they are not very detailed in places or I might be missing/misunderstanding something. Edit 2: While I figured ...
1 vote
0 answers
871 views
How to display image from CUDA / What API to port my project to for real-time raytracing?
I started a raytracing project using the python numba library which provides a just-in-time compiler for CUDA kernels. When the scene is rendered my result is a 1920 x 1080 x 3 RGB array in GPU memory....
- 71
2 votes
1 answer
528 views
Strange sphere reflection effects in ray tracing project. I cannot figure out what causes them
I started a ray tracing project in Python. I made a CUDA kernel and a trace() function for each pixel. It works perfectly when calculating intersections, shadows and lambert shading, however when I ...
- 71
1 vote
0 answers
83 views
Why cache working set per multiprocessor for texture memory in Nvidia has a variable size?
I saw it here https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#features-and-technical-specifications__technical-specifications-per-compute-capability , I don't know if it also happens ...
- 11
4 votes
2 answers
1k views
Can gl_WorkGroupSize be set outside the compute shader?
I am migrating code from CUDA to OpenGL compute shaders. I could not find a way to set the local workgroup size outside the shader; is there a way to set gl_WorkGroupSize in the calling program like ...
- 319
3 votes
1 answer
148 views
How to avoid slowdown with 25-30 students running simple GPU kernels on 4 GeForce GTX 650 Ti s?
So I'm teaching crash-course in CUDA that teaches students how to write good GPU code (CUDA 7.5 in this case). They kernels they will be running will do matrix multiply on 2048x2048 floating point ...
- 33