• Advertisement
  • Popular Tags

  • Popular Now

  • Advertisement
  • Similar Content

    • By racarate
      Hey everybody!
      I am trying to replicate all these cool on-screen debug visuals I see in all the SIGGRAPH and GDC talks, but I really don't know where to start.  The only resource I know of is almost 16 years old:
      http://number-none.com/product/Interactive Profiling, Part 1/index.html
      Does anybody have a more up-to-date reference?  Do people use minimal UI libraries like Dear ImgGui?  Also, If I am profiling OpenGL ES 3.0 (which doesn't have timer queries) is there really anything I can do to measure performance GPU-wise?  Or should I just chart CPU-side frame time?  I feel like this is something people re-invent for every game there has gotta be a tutorial out there... right?
       
       
    • By Achivai
      Hey, I am semi-new to 3d-programming and I've hit a snag. I have one object, let's call it Object A. This object has a long int array of 3d xyz-positions stored in it's vbo as an instanced attribute. I am using these numbers to instance object A a couple of thousand times. So far so good. 
      Now I've hit a point where I want to remove one of these instances of object A while the game is running, but I'm not quite sure how to go about it. At first my thought was to update the instanced attribute of Object A and change the positions to some dummy number that I could catch in the vertex shader and then decide there whether to draw the instance of Object A or not, but I think that would be expensive to do while the game is running, considering that it might have to be done several times every frame in some cases. 
      I'm not sure how to proceed, anyone have any tips?
    • By fleissi
      Hey guys!

      I'm new here and I recently started developing my own rendering engine. It's open source, based on OpenGL/DirectX and C++.
      The full source code is hosted on github:
      https://github.com/fleissna/flyEngine

      I would appreciate if people with experience in game development / engine desgin could take a look at my source code. I'm looking for honest, constructive criticism on how to improve the engine.
      I'm currently writing my master's thesis in computer science and in the recent year I've gone through all the basics about graphics programming, learned DirectX and OpenGL, read some articles on Nvidia GPU Gems, read books and integrated some of this stuff step by step into the engine.

      I know about the basics, but I feel like there is some missing link that I didn't get yet to merge all those little pieces together.

      Features I have so far:
      - Dynamic shader generation based on material properties
      - Dynamic sorting of meshes to be renderd based on shader and material
      - Rendering large amounts of static meshes
      - Hierarchical culling (detail + view frustum)
      - Limited support for dynamic (i.e. moving) meshes
      - Normal, Parallax and Relief Mapping implementations
      - Wind animations based on vertex displacement
      - A very basic integration of the Bullet physics engine
      - Procedural Grass generation
      - Some post processing effects (Depth of Field, Light Volumes, Screen Space Reflections, God Rays)
      - Caching mechanisms for textures, shaders, materials and meshes

      Features I would like to have:
      - Global illumination methods
      - Scalable physics
      - Occlusion culling
      - A nice procedural terrain generator
      - Scripting
      - Level Editing
      - Sound system
      - Optimization techniques

      Books I have so far:
      - Real-Time Rendering Third Edition
      - 3D Game Programming with DirectX 11
      - Vulkan Cookbook (not started yet)

      I hope you guys can take a look at my source code and if you're really motivated, feel free to contribute :-)
      There are some videos on youtube that demonstrate some of the features:
      Procedural grass on the GPU
      Procedural Terrain Engine
      Quadtree detail and view frustum culling

      The long term goal is to turn this into a commercial game engine. I'm aware that this is a very ambitious goal, but I'm sure it's possible if you work hard for it.

      Bye,

      Phil
    • By tj8146
      I have attached my project in a .zip file if you wish to run it for yourself.
      I am making a simple 2d top-down game and I am trying to run my code to see if my window creation is working and to see if my timer is also working with it. Every time I run it though I get errors. And when I fix those errors, more come, then the same errors keep appearing. I end up just going round in circles.  Is there anyone who could help with this? 
       
      Errors when I build my code:
      1>Renderer.cpp 1>c:\users\documents\opengl\game\game\renderer.h(15): error C2039: 'string': is not a member of 'std' 1>c:\program files (x86)\windows kits\10\include\10.0.16299.0\ucrt\stddef.h(18): note: see declaration of 'std' 1>c:\users\documents\opengl\game\game\renderer.h(15): error C2061: syntax error: identifier 'string' 1>c:\users\documents\opengl\game\game\renderer.cpp(28): error C2511: 'bool Game::Rendering::initialize(int,int,bool,std::string)': overloaded member function not found in 'Game::Rendering' 1>c:\users\documents\opengl\game\game\renderer.h(9): note: see declaration of 'Game::Rendering' 1>c:\users\documents\opengl\game\game\renderer.cpp(35): error C2597: illegal reference to non-static member 'Game::Rendering::window' 1>c:\users\documents\opengl\game\game\renderer.cpp(36): error C2597: illegal reference to non-static member 'Game::Rendering::window' 1>c:\users\documents\opengl\game\game\renderer.cpp(43): error C2597: illegal reference to non-static member 'Game::Rendering::window' 1>Done building project "Game.vcxproj" -- FAILED. ========== Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped ==========  
       
      Renderer.cpp
      #include <GL/glew.h> #include <GLFW/glfw3.h> #include "Renderer.h" #include "Timer.h" #include <iostream> namespace Game { GLFWwindow* window; /* Initialize the library */ Rendering::Rendering() { mClock = new Clock; } Rendering::~Rendering() { shutdown(); } bool Rendering::initialize(uint width, uint height, bool fullscreen, std::string window_title) { if (!glfwInit()) { return -1; } /* Create a windowed mode window and its OpenGL context */ window = glfwCreateWindow(640, 480, "Hello World", NULL, NULL); if (!window) { glfwTerminate(); return -1; } /* Make the window's context current */ glfwMakeContextCurrent(window); glViewport(0, 0, (GLsizei)width, (GLsizei)height); glOrtho(0, (GLsizei)width, (GLsizei)height, 0, 1, -1); glMatrixMode(GL_PROJECTION); glLoadIdentity(); glfwSwapInterval(1); glEnable(GL_SMOOTH); glEnable(GL_DEPTH_TEST); glEnable(GL_BLEND); glDepthFunc(GL_LEQUAL); glHint(GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST); glEnable(GL_TEXTURE_2D); glLoadIdentity(); return true; } bool Rendering::render() { /* Loop until the user closes the window */ if (!glfwWindowShouldClose(window)) return false; /* Render here */ mClock->reset(); glfwPollEvents(); if (mClock->step()) { glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); glfwSwapBuffers(window); mClock->update(); } return true; } void Rendering::shutdown() { glfwDestroyWindow(window); glfwTerminate(); } GLFWwindow* Rendering::getCurrentWindow() { return window; } } Renderer.h
      #pragma once namespace Game { class Clock; class Rendering { public: Rendering(); ~Rendering(); bool initialize(uint width, uint height, bool fullscreen, std::string window_title = "Rendering window"); void shutdown(); bool render(); GLFWwindow* getCurrentWindow(); private: GLFWwindow * window; Clock* mClock; }; } Timer.cpp
      #include <GL/glew.h> #include <GLFW/glfw3.h> #include <time.h> #include "Timer.h" namespace Game { Clock::Clock() : mTicksPerSecond(50), mSkipTics(1000 / mTicksPerSecond), mMaxFrameSkip(10), mLoops(0) { mLastTick = tick(); } Clock::~Clock() { } bool Clock::step() { if (tick() > mLastTick && mLoops < mMaxFrameSkip) return true; return false; } void Clock::reset() { mLoops = 0; } void Clock::update() { mLastTick += mSkipTics; mLoops++; } clock_t Clock::tick() { return clock(); } } TImer.h
      #pragma once #include "Common.h" namespace Game { class Clock { public: Clock(); ~Clock(); void update(); bool step(); void reset(); clock_t tick(); private: uint mTicksPerSecond; ufloat mSkipTics; uint mMaxFrameSkip; uint mLoops; uint mLastTick; }; } Common.h
      #pragma once #include <cstdio> #include <cstdlib> #include <ctime> #include <cstring> #include <cmath> #include <iostream> namespace Game { typedef unsigned char uchar; typedef unsigned short ushort; typedef unsigned int uint; typedef unsigned long ulong; typedef float ufloat; }  
      Game.zip
    • By lxjk
      Hi guys,
      There are many ways to do light culling in tile-based shading. I've been playing with this idea for a while, and just want to throw it out there.
      Because tile frustums are general small compared to light radius, I tried using cone test to reduce false positives introduced by commonly used sphere-frustum test.
      On top of that, I use distance to camera rather than depth for near/far test (aka. sliced by spheres).
      This method can be naturally extended to clustered light culling as well.
      The following image shows the general ideas

       
      Performance-wise I get around 15% improvement over sphere-frustum test. You can also see how a single light performs as the following: from left to right (1) standard rendering of a point light; then tiles passed the test of (2) sphere-frustum test; (3) cone test; (4) spherical-sliced cone test
       

       
      I put the details in my blog post (https://lxjk.github.io/2018/03/25/Improve-Tile-based-Light-Culling-with-Spherical-sliced-Cone.html), GLSL source code included!
       
      Eric
  • Advertisement
  • Advertisement
Sign in to follow this  

OpenGL Naivete Challenged: Interleaving arrays causes slow performance?

This topic is 1913 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Ok, so I was thinking about how to speed up my opengl application, and remembered hearing something about interleaving array data. To my simple mind it seemed that when I pass attributes to a shader in separate buffer objects some kind of jumping around in the video cards' memory would occur, so if I interleaved the data for every vertex then this should provide the benefit of locality of reference like in a cache, and things should speed up.

 

However after some experimenting it cause a decline of 2 frames per second in my overall speed. So WHY does this happen? I'm clearly drinking from the wrong bottle of GL goodness. 

Share this post


Link to post
Share on other sites
Advertisement

2 frames per second

2 frames of what ? FPS is not the best measurement, 2 fps of 4 fps is much more than 2 fps of 10000 fps.

 

From what you describes I would say, that vertex perfomance is not your bottleneck and a drop of 2 fps of X is not a proof of interlaeving arrays are slower than not interleaved ones.

Edited by Ashaman73

Share this post


Link to post
Share on other sites

so if I interleaved the data for every vertex then this should provide the benefit of locality of reference like in a cache, and things should speed up.

Did you make sure that each vertex is a power-of-2 size? If your vertex is an odd size it will cross multiple cache lines - 16, 32 or 64 bytes are typically good sizes.

Share this post


Link to post
Share on other sites

As a general rule interleaving should be faster, but there is one case in which it may be slower, and that's if the per-vertex pipeline is either implemented in or being emulated by software.  In that case splitting off the position component of your arrays, and padding it to 4 floats to better allow for SIMD optimizations in your driver, may be the optimal solution.

 

You haven't really given much info about the kind of workload you're running or what kind of hardware you've got, but some circumstances that can cause per-vertex to run in software would include having older graphics hardware that doesn't support hardware T&L (some Intel integrateds from about 5/6 years or more ago would fall under this category), or exceeding a hardware limit/trying to do something that's allowed by the GL spec but is not actually supported in hardware (that may include using GL_UNSIGNED_INT for your GL_ELEMENT_ARRAY_BUFFER on older hardware - again, you have to go back at least 6 years for this - or using GL_UNSIGNED_BYTE on any hardware).  If any of this sounds familiar you may have just identified your cause.

Share this post


Link to post
Share on other sites

so if I interleaved the data for every vertex then this should provide the benefit of locality of reference like in a cache, and things should speed up.

Did you make sure that each vertex is a power-of-2 size? If your vertex is an odd size it will cross multiple cache lines - 16, 32 or 64 bytes are typically good sizes.

 

Aha! I had very odd vertex sizes, 12 bytes for a position, 12 bytes for a normal, 8 bytes for a texture or  just 12 bytes for a position, 8 bytes for a texture. So I will pad the vertices, textures and normals out to 16 bytes each but make each stride between a whole bunch of vertex data 64 bytes as 48 bytes isn't a power of two. Is this a good strategy?

 

Are there any useful links or articles concerning this kind of thing and the strategies you should use?

 

 

 

As a general rule interleaving should be faster, but there is one case in which it may be slower, and that's if the per-vertex pipeline is either implemented in or being emulated by software.  In that case splitting off the position component of your arrays, and padding it to 4 floats to better allow for SIMD optimizations in your driver, may be the optimal solution.

 

You haven't really given much info about the kind of workload you're running or what kind of hardware you've got, but some circumstances that can cause per-vertex to run in software would include having older graphics hardware that doesn't support hardware T&L (some Intel integrateds from about 5/6 years or more ago would fall under this category), or exceeding a hardware limit/trying to do something that's allowed by the GL spec but is not actually supported in hardware (that may include using GL_UNSIGNED_INT for your GL_ELEMENT_ARRAY_BUFFER on older hardware - again, you have to go back at least 6 years for this - or using GL_UNSIGNED_BYTE on any hardware).  If any of this sounds familiar you may have just identified your cause.

 

I'm using an AMD HD Radeon 4200 card. Is this a terrible Graphics card?

Share this post


Link to post
Share on other sites

Did you make sure that each vertex is a power-of-2 size? If your vertex is an odd size it will cross multiple cache lines - 16, 32 or 64 bytes are typically good sizes.

 
Aha! I had very odd vertex sizes, 12 bytes for a position, 12 bytes for a normal, 8 bytes for a texture or  just 12 bytes for a position, 8 bytes for a texture. So I will pad the vertices, textures and normals out to 16 bytes each but make each stride between a whole bunch of vertex data 64 bytes as 48 bytes isn't a power of two. Is this a good strategy?

Actually it isn’t about it being a power of 2, but a multiple of the cache size, which is usually 16 or 32. MSDN quotes 32 explicitly for certain cases of vertex buffers but my testing reveals this to improve the performance in all cases, not just what they list.  Of course, their documentation is for Direct3D, but cache issues and friends are a universal issue.

Performance Optimizations

 

Also, we are not talking about padding each vertex-buffer element.  12-byte normals and positions (etc.) are completely normal.

This padding is between each vertex in the buffer, not each element of each vertex.

Normally once you interleave position, normals, UV’s, tangents, and bitangents, you have 56-byte vertices.  This should be padded out to 64.

 

 

L. Spiro

Edited by L. Spiro

Share this post


Link to post
Share on other sites

Actually it isn’t about it being a power of 2, but a multiple of the cache size, which is usually 16 or 32.

Sure. And a power of two will always be an even multiple/divisor of the cache size.

It's accurate enough for a rule of thumb *shrugs*

Aha! I had very odd vertex sizes, 12 bytes for a position, 12 bytes for a normal, 8 bytes for a texture

12 + 12 + 8 = 32 bytes, which is just the right size for cache efficiency.

If you weren't interleaving these you'd have to be more careful about the individual sizes, but for an interleaved vertex it's the structure as a whole that matters. Edited by swiftcoder

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement