7
votes
Accepted
How much precision (half, float, double, etc) is enough for a Color class?
Colors shown on your display or saved to standard image file formats use 8 bits per component. So to store these colors it suffices to use four bytes (unsigned char)...
- 25.2k
4
votes
Accepted
What happens with the framebuffer after the fragment shader is done?
It entirely depends what hardware and OS you're on. It's up to the integration between the windowing system and GL. In the simple case where one application is rendering full-screen directly, the ...
- 6,870
3
votes
Accepted
Is it Possible to send data directly to the GPU via a custom hardware interface?
On a desktop (or laptop) the GPU is connected to the PCIe bus. This gives it direct access to the physical RAM.
If you map the buffer you want to fill and then read a file into that mapped memory. ...
- 6,000
3
votes
Accepted
What data is passed from the CPU to the GPU each frame?
How much data is passed from the CPU to the GPU per frame depends on the engine you are using and its needs. Usually, you want to keep it minimal to avoid that the GPU needs to wait for data because ...
- 1,251
3
votes
Accepted
What do the OpenGL version strings from glxinfo actually mean?
When creating a OpenGL context you have to specify a profile mask, which is a way of telling the driver whether you want to use compatibility ("legacy") features or not.
Drivers are required to ...
- 237
3
votes
Wouldn't it be better if the vertex shaders and Z-Buffer were on the CPU instead of the GPU?
… you could do the calculation of the polygon world directly on the CPU die and filter out the triangles that shouldn't be shown using the Z-buffer and then send only the triangles that should be ...
- 1,276
3
votes
What were the benefits of quads used by Nvidia's NV1 over tri based rendering?
There seems to be a lot of confusion around the NV1, so I tried to find a few actual sources.
It looks like NVidia and SGS-Thomson entered a partnership [#1] [#2] for a "Multimedia Accelerator&...
- 138
2
votes
Accepted
How can I test my GLSL code on different hardware?
I highly recommend getting at least 1 card from one of the other vendors like AMD. It doesn't have to be fancy, the cheapest thing you can get off ebay will help. Just make sure it supports the same ...
- 3,354
2
votes
Accepted
On modern GPUs, what hardware do vertex/pixel shaders have access to that compute shaders don't?
what hardware do standard graphics pipelines have access to which general purpose GPU programming APIs dont give access to?
Every part of the graphics pipeline and architecture that isn't a shader. ...
- 9,932
2
votes
Accepted
Compute shader workgroups execution and size
Since you mentioned CUs and SIMDs, I wrote this mainly from the perspective of AMD's GCN architecture, but broadly speaking the answers apply to other GPU architectures as well (except for #2 which is ...
- 25.2k
1
vote
Accepted
What is the oldest hardware able to feasibly support Vulkan?
Khronos publishes a list of Vulkan compliant hardware here. The list also shows compliance to different flavors of OpenGL. I don't know all the card models, but for NVIDIA at least the GeForce GT 630 ...
- 830
1
vote
Precision problem on AMD
In HLSL, you can use the built-in fmod function for this.
I'm not sure what's going wrong with your implementation, though. It looks mathematically correct. It's ...
- 25.2k
1
vote
What data is passed from the CPU to the GPU each frame?
The short answer to this question is: everything
On any give frame almost any kind of data my be needed to render that new and wonderful object that just popped into view. It's shader needs to be ...
- 3,354
1
vote
Rasterization: How does hardware disambiguate identical Z values at silhouette edges?
If you are taking about this occurring within a single draw call for a single object then you are talking about rasterization not Z fighting. So the real question is, "how are two triangles ...
- 3,354
1
vote
Accepted
How does the OS + hardware support multiple screen resolutions and multi-display?
Any hardware in the graphics card is fixed.
No, the storage of a graphics card is fixed. How that storage gets used, particularly on any semi-modern graphics card, is flexible.
Essentially, what a ...
- 9,932
1
vote
FLOP - exact definition
Just to add to whychmaster's reply, in my experience, on a GPU, a floating-point operation, when used in quoting FLOPS (floating-point operations per second) benchmark figures will refer to Addition (...
- 4,306
1
vote
Why did AMD's putting an SSD on a GPU board reduce latency so much?
Answering from a somewhat layperson's perspective (I'm not a hardware expert, but I have built my own computer), I'll analogize, using the situation of getting some ingredients for cooking food.
...
- 153
1
vote
MSI GTX 950 2GD5T triangles per second
In my tests a GTX 1050 can do ~1B triangles with glDrawElementsInstanced(GL_TRIANGLES, .... That's roughly 2/3 of chip clock. Arguably ...
- 111
Only top scored, non community-wiki answers of a minimum length are eligible