Pagur said:
But this is clearly not the number provided by AMD which is just 304.8 GTexel/s (1905 MHz x 160 TMU).

A RX 5700 has memory bandwidth of 448 GB/s, and assuming a texel is one byte, this number matches better.

Pagur said:
it seems legit to compute the texture fill rate as follow: Texture Rate…

Hi,

I'm working on a game in UE4 that uses a compute shader to perform 3D rotations on 2D images. The final rotated image is drawn to a render target after the calculations are done.

Problem: on certain hardware, we're seeing “striping” occur, where rows of pixels are missing from the final image (ex…

What ^ @joej said. But I'll add a personal axiom that if it's atomic and it's not an index or a counter … you probably need to think again. It applies to all atomics but Compared (min/max) atomics (like an atomic ray-hit distance) are the more obvious example of how atomics can become problems in h…

Why do you sort the (seemingly opaque) tiles? Are you sure this improves performance instead of just rendering them naively?

I build my terrain mesh on the CPU but in chunks. Maybe you could separate it into chunks as well, assuming updating the terrain (CPU → GPU) really is an issue with performanc…

This is my Sharpview metaverse viewer, built on top of the Rust Rend3/WGPU/Vulkan stack.

Video: https://video.hardlimit.com/w/tp9mLAQoHaFR32YAVKVDrz​​

Part of my ongoing effort to prove that metaverses full of complex user-create…

Hello everyone,

I'm working on a rasterizer that renders triangles in 2d space, all additively (no depth checking). I'm looking for tips on performance optimization of all types. I don't have a lot of experience with OpenCL/memory management etc so I'm very happy to learn about all the ways to impro…

@frob Yes.I am trying to use PIX to find where the problem is. But,you know,I am a freshman in DirectX. So the process is a bit difficult for me.Could you please tell me some good tutorials about DirectX12 or PIX？Thank you again for this!

memory-map the buffer (glMapBuffer) …

Ooooh didn't know that one, will take a look at it!

@Aressera thanks a lot for answering, your explanations really helped me out!

@foxo You need to create a bitmap and temporary device context (DC) that's compatible with your window. You can get a pointer to the bitmap pixels using GetDIBits. All drawing should be done using the pointer. You then copy the pointer back to the bitmap/dc using SetDIBits. Last step is to blit the…

foxo said:

I also realised that gpu drivers and extremely low level gpu programming are some kind a ‘secret’ knoladge that nobody know how to do except the elit at vulkan and other specialised company that won't tell us there ‘secrets’. that's why I am asking : How to draw from gpu to the screen or …

Yes, the pre-allocated mesh structure must live past a single frame.

Just like textures and shaders – you don't want to re-upload textures, or re-compile shaders, every frame.

If you use a texture array where each “channel ” (diffuse/spec/normal/emissive) has an associated index for each material, you could use a special index to indicate that it should use a default value instead of sampling an actual texture. Then in the shader you could just branch on checking for tha…

Using profiling to inform coding decisions; a real life example.  

To start off, I added the ability for ships to respawn. Giving some forgiveness if the player dies.

That works by adding a spawn component to the entity that needs to respawn fighter ships.

When respawning, you need to know how lo…

Few (if any) GPU-driven pipelines can draw without any state changes, which means you can't generally get away with a single call to DrawIndexedIndirect. Something close might be possible if you bucket all your materials into a small number of “uber” shaders (e.g. “opaque”, “transparent”, “cutout”,…

There is no real technical difference between gfx ‘shaders’ and GPGPU 'compute', both use the same cores and will equally benefit from higher frequency. 
Dedicated gfx HW is mainly the non programmable stuff like ROPs or TMUs, and now RT, and those units usually also depend on the same core clo…

When a DX11 application executes Map/Unmap with D3D11_MAP_WRITE_NO_OVERWRITE, is there a way to efficiently know which region of the provided buffer was modified?

I hooked all DX11 calls of “Heroes of the storm” game and made it playable on low-end machines.
https://www.reddit.com/…

Why do you even need DirectX 9? Now any video card already supports 11. And it is much easier to learn thanks to more complete documentation and has more features.

Just a wild guess, but a game written in DirectX8 will probably not make much use of multithreading, or does it? If not, I assume you have a PC with 8 threads? If so, the ~12% CPU-usage would mean your game is CPU-locked on one thread, meaning the GPU might be just idling for >⅔rds of the time w…

All games so far only use raytracing as an optional part of the pipeline, other than an obscure game called Claybook and a 3DMark benchmark.

Rasterization is always used for primary visibility in all those other games.

GPU is a state machine, only calls that goes operating on it- are the state changing ones, dx11 contains at least a few utility functions which are not GPU calls. Or maybe none, who knows, I do not contain all of it and I have used only GPU calls of the API.

well done and thank you for sharing your solution ?

So you know your game is heavy on the GPU and you're ready to optimize it.

But wait, where exactly do YOU start optimizing? Do you know which parts of your scene are causing your GPU performance bottlenecks?

Well, I have a tool to help you with Unity GPU Performance profiling: RenderDoc.

Quick Navigat…

In this post, I'll show you how to know if your game performance is suffering from too many Unity Draw Calls and what to do about it. You'll be able to answer:

• What is a draw call and what will they do to your players?
• How do you know if you have too many draw calls?
• How can you reduce these dreadful …

Is your game rendering more geometry than your GPU can handle? Here, let me show you a few steps to check if your geometry is causing a performance bottleneck on your players' GPU.

In this blog post, I'll share with you:

• What geometry complexity is and how it affects your players.
• How to check if geom…

AMD has released the "Vega" 7nm Instruction Set Architecture in PDF form. They have been releasing GPU ISA documentation for some time now, and now they've added the 7nm architecture to the list.

In the ISA documentation you will find:

• Chapter 1 begins the document begins with an overview of the AMD …