• Advertisement

OpenGL Low framerate during first seconds

Recommended Posts

Hi gamedev !

I'm facing a really strange performance issue occurring during the first seconds of my app.
When launching the app, three scenarios:

  • Some erratic frames at the very beginning, then 70+fps.
  • Some erratic frames at the very beginning, then ~12fps, then, after a few seconds, BAM, 70+fps.
  • Some erratic frames at the very beginning, then stuck at ~12fps.

Camera is static, so is the scene.

I used nsight to try figure out what's happening, especially in case 2 where I can see what are the differences.
When it's rendering at ~12fps, some glDrawRangeElementsBaseVertex/glDrawArrays calls takes a lot of time, some SwapBuffers calls too, and after switching to "cruise speed", those calls are taking a more moderate amount of time:

Here is a capture of a nsight session timeline:


1->erratic framerate at the beginning


3->stable and high framerate

Memory usage is the same in all cases.
CPU Usage is a bit higher when stuck at ~12fps, but nothing remarkable.

My scene is really simple (only a few cubes), but I use a lot of post process (SSAO, SSR). Reducing the framebuffer size to a smaller one doesn't change anything (although top speed is faster).
I also tried to disable all my post process, issue is still there but case 1 is the most common.

it is really similar to what's described in this post, although I'm not using vsync at all (double checked code, and also forced vsync off with the nvidia control panel) nor SDL.

I also tried to put some glFlush()/glFinish() here and there in my code, without success.

A few specifications:
OS: Windows 7
CPU: Core I7 6700
GPU: Nvidia 560Ti
Drivers: 385.41
I'm stuck with this issue for months now, it's driving me crazysick.gif

Thanks for any help/advice/question/test to make

Edited by CrashyCartman

Share this post

Link to post
Share on other sites

Throwing a dart in the dark here...

When you launch your app, everything is cold. Driver has to setup who knows what for your application in the background, compile your shaders, pass stuff to the GPU, etc. Once all of this stabilizes (say, you've used all your shaders at least once for actual drawing), everything runs as expected.

Drivers do a lot of lazy loading, specially for OpenGL where there are tons of paths to do the same thing and they might require slightly different things. You might ask the driver to compile a shader, and bind a bunch of state, but the driver might only actually do that work when you actually draw with any of them set, until that happens, the driver can just "pretend" it did what you told it to do. It saves up work if you say, change the raster state 3 or 4 times until you actually draw with it. Driver makers prefer to assume you're stupid rather than taking the backlash of XYZ big game that doesn't works as fast as the competition.

Also if you're using any kind of "managed language" (something that runs on top of the CLR or a JVM for instance), those have to do a lot of setup work too, usually concurrently with your application (say, HotSpot, a JVM, can run your code in interpreted mode, while it is compiling it at the same time, then later replace it with the compiled code when it finishes).

This is also true of native languages to a lesser extent for different reasons, big exe/dll/so/dylib/etc might take a while to load and the important bits might take a while until  they hit the CPU caches.

Share this post

Link to post
Share on other sites

Thanks for the answer:)

I know driver must be busy after resource loading, and could accept having some long frames at start, but in this case, 33% of time I've got a game completely unplayable.

I also made tests removing all of my threads, just to see if it was a concurrency issue, and unfortunately it didn't change anything.

Everything is in c++, no managed code.

I may add that if I run the exe by doing "Nsight/Start Graphics Debugging", everything is ok, of course I have some hiccups at start but then, game runs smooth every time.


Share this post

Link to post
Share on other sites

Ah right, I misunderstood the problem. Just in case, hows GPU memory usage? Most 560 have 1Gb, you might be threading in a line where sometimes you're running out of GPU memory and sometimes you don't, for whatever reason.

You can check it in a nice graph with something like MSI Afterburner or GPU-Z.

Check system memory too just in case.

Edited by TheChubu

Share this post

Link to post
Share on other sites

Alright here are the results of my quick benchmarks made using GPU-Z to track gpu memory.

During loading and the first erratic frames, gpu memory is full (1000MB).

After that, it falls down to around 800 MB, framerate is stable but low.

And if after a few seconds framerate increases, GPU memory goes to around 950MB, if not, it stays at 800MB/low fps forever.

So there is a real correlation between both memory usage & performance.

I tried changing the format of the few FLOAT16 RGBA  FBOs I use to RGBA8 in order to reduce size, but it didn't change anything to memory usage, which is strange.

System memory shows the same relationship between memory & framerate.


1: first run, app runs fine at 70 fps

2: second run, app is stuck at 12 fps.

3: third run, same as 1

As you can see, the difference is small (100 MB) but noticeable.

Thanks for trying to help me :)

Share this post

Link to post
Share on other sites

Are you running in fullscreen at 4k or something? Say, on my Windows 7 desktop I get around 100Mb of GPU memory used constantly, the rest is available. So if you have anything else running that hogs VRAM, close it (games, videos, chrome, whatever). See how much your application is actually taking. Try running it in a 800x600 window.

Share this post

Link to post
Share on other sites

Running in 720p, and yes, I find this crazy to have so much vram used, should have tracked that before. I use only a few small textures for the moment, no reason to have so much memory used.

Thanks for your help, I'll check what's wront and post the solution here :)

Share this post

Link to post
Share on other sites

Alright, here are the various causes :)

  1. I had some extra large (8k) textures I used to generate small sdf font textures, but never removed them after sdf generation.
  2. My shadow maps are packed into a single big GL_R32F atlas, and the depth buffer of this FBO is as large as the atlas, so it's a huge amount of memory. Correct me if I'm wrong, but I think if I use GL_DEPTH_COMPONENT32F as a format I'll reduce memory used by 2.

Thanks again for pointing me in the right direction.

Share this post

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Advertisement
  • Advertisement
  • Popular Tags

  • Advertisement
  • Popular Now

  • Similar Content

    • By LifeArtist
      Good Evening,
      I want to make a 2D game which involves displaying some debug information. Especially for collision, enemy sights and so on ...
      First of I was thinking about all those shapes which I need will need for debugging purposes: circles, rectangles, lines, polygons.
      I am really stucked right now because of the fundamental question:
      Where do I store my vertices positions for each line (object)? Currently I am not using a model matrix because I am using orthographic projection and set the final position within the VBO. That means that if I add a new line I would have to expand the "points" array and re-upload (recall glBufferData) it every time. The other method would be to use a model matrix and a fixed vbo for a line but it would be also messy to exactly create a line from (0,0) to (100,20) calculating the rotation and scale to make it fit.
      If I proceed with option 1 "updating the array each frame" I was thinking of having 4 draw calls every frame for the lines vao, polygons vao and so on. 
      In addition to that I am planning to use some sort of ECS based architecture. So the other question would be:
      Should I treat those debug objects as entities/components?
      For me it would make sense to treat them as entities but that's creates a new issue with the previous array approach because it would have for example a transform and render component. A special render component for debug objects (no texture etc) ... For me the transform component is also just a matrix but how would I then define a line?
      Treating them as components would'nt be a good idea in my eyes because then I would always need an entity. Well entity is just an id !? So maybe its a component?
    • By QQemka
      Hello. I am coding a small thingy in my spare time. All i want to achieve is to load a heightmap (as the lowest possible walking terrain), some static meshes (elements of the environment) and a dynamic character (meaning i can move, collide with heightmap/static meshes and hold a varying item in a hand ). Got a bunch of questions, or rather problems i can't find solution to myself. Nearly all are deal with graphics/gpu, not the coding part. My c++ is on high enough level.
      Let's go:
      Heightmap - i obviously want it to be textured, size is hardcoded to 256x256 squares. I can't have one huge texture stretched over entire terrain cause every pixel would be enormous. Thats why i decided to use 2 specified textures. First will be a tileset consisting of 16 square tiles (u v range from 0 to 0.25 for first tile and so on) and second a 256x256 buffer with 0-15 value representing index of the tile from tileset for every heigtmap square. Problem is, how do i blend the edges nicely and make some computationally cheap changes so its not obvious there are only 16 tiles? Is it possible to generate such terrain with some existing program?
      Collisions - i want to use bounding sphere and aabb. But should i store them for a model or entity instance? Meaning i have 20 same trees spawned using the same tree model, but every entity got its own transformation (position, scale etc). Storing collision component per instance grats faster access + is precalculated and transformed (takes additional memory, but who cares?), so i stick with this, right? What should i do if object is dynamically rotated? The aabb is no longer aligned and calculating per vertex min/max everytime object rotates/scales is pretty expensive, right?
      Drawing aabb - problem similar to above (storing aabb data per instance or model). This time in my opinion per model is enough since every instance also does not have own vertex buffer but uses the shared one (so 20 trees share reference to one tree model). So rendering aabb is about taking the model's aabb, transforming with instance matrix and voila. What about aabb vertex buffer (this is more of a cosmetic question, just curious, bumped onto it in time of writing this). Is it better to make it as 8 points and index buffer (12 lines), or only 2 vertices with min/max x/y/z and having the shaders dynamically generate 6 other vertices and draw the box? Or maybe there should be just ONE 1x1x1 cube box template moved/scaled per entity?
      What if one model got a diffuse texture and a normal map, and other has only diffuse? Should i pass some bool flag to shader with that info, or just assume that my game supports only diffuse maps without fancy stuff?
      There were several more but i forgot/solved them at time of writing
      Thanks in advance
    • By RenanRR
      Hi All,
      I'm reading the tutorials from learnOpengl site (nice site) and I'm having a question on the camera (https://learnopengl.com/Getting-started/Camera).
      I always saw the camera being manipulated with the lookat, but in tutorial I saw the camera being changed through the MVP arrays, which do not seem to be camera, but rather the scene that changes:
      Vertex Shader:
      #version 330 core layout (location = 0) in vec3 aPos; layout (location = 1) in vec2 aTexCoord; out vec2 TexCoord; uniform mat4 model; uniform mat4 view; uniform mat4 projection; void main() { gl_Position = projection * view * model * vec4(aPos, 1.0f); TexCoord = vec2(aTexCoord.x, aTexCoord.y); } then, the matrix manipulated:
      ..... glm::mat4 projection = glm::perspective(glm::radians(fov), (float)SCR_WIDTH / (float)SCR_HEIGHT, 0.1f, 100.0f); ourShader.setMat4("projection", projection); .... glm::mat4 view = glm::lookAt(cameraPos, cameraPos + cameraFront, cameraUp); ourShader.setMat4("view", view); .... model = glm::rotate(model, glm::radians(angle), glm::vec3(1.0f, 0.3f, 0.5f)); ourShader.setMat4("model", model);  
      So, some doubts:
      - Why use it like that?
      - Is it okay to manipulate the camera that way?
      -in this way, are not the vertex's positions that changes instead of the camera?
      - I need to pass MVP to all shaders of object in my scenes ?
      What it seems, is that the camera stands still and the scenery that changes...
      it's right?
      Thank you
    • By dpadam450
      Sampling a floating point texture where the alpha channel holds 4-bytes of packed data into the float. I don't know how to cast the raw memory to treat it as an integer so I can perform bit-shifting operations.

      int rgbValue = int(textureSample.w);//4 bytes of data packed as color
      // algorithm might not be correct and endianness might need switching.
      vec3 extractedData = vec3(  rgbValue & 0xFF000000,  (rgbValue << 8) & 0xFF000000, (rgbValue << 16) & 0xFF000000);
      extractedData /= 255.0f;
    • By Devashish Khandelwal
      While writing a simple renderer using OpenGL, I faced an issue with the glGetUniformLocation function. For some reason, the location is coming to be -1.
      Anyone has any idea .. what should I do?
  • Advertisement