Sign in to follow this  

OpenGL Performance issues rendering triangles vs tri-strips

This topic is 3737 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

Hello all, I've just converted a class that generated a triangle-strip mesh of a skydome to generating it with triangles. I did this as I've read a couple of articles that state that rendering triangles is slightly faster than strips because the GPU is able to take advantage of its fast vertex cache. Also, numerous posts here on GameDev by many gurus state just the same. However, after converting the class I tested it with the old tri-strip (a) and triangle (b) skydome meshes hoping I would get an increase in FPS (if only small.) Results: a) Tri-strip
================================================
Viewport: (0, 0, 1024, 768)
Run time: 36470ms , ~36s
Total frames: 29984
Highest frame rate: 865
Lowest frame rate: 759
Average frame rate: 832
b) Triangle mesh
================================================
Viewport: (0, 0, 1024, 768)
Run time: 90693ms , ~90s
Total frames: 67752
Highest frame rate: 780
Lowest frame rate: 692
Average frame rate: 752
The two test programs are release builds, were ran in 1024x768 full-screen res and render about 20000 triangles, though the skydome mesh alone consists only of 5180 triangles (5184 tri-strip elements in a; 15540 vertices in b.) I let test b) run for longer because I couldn't believe the (significant) drop in FPS and was hoping for some miracle to happen... My graphics card is an ATI Mobility Radeon x700 (128MB, PCIe.) What could be the reason to the drop in FPS? <edit> The new skydome generating algorithm is in essence the same as it was before when it generated a tri-strip mesh. The only differences now is that the vertex buffer is larger so as to accomodate all the triangle vertices of the skydome ((vertices-2)*3) and also each vertice is stored at every 3rd position in the vertex buffer after the first 3 elements (vertexbuffer[n*3]=triangleVertice, n>3, 1 being the lowest index.) Then I iterate through the vertex buffer to finalize the triangles by using OpenGL's rules when rendering triangle_strips {odd=(n,n+1,n+2);even=(n+1,n,n+2)}. All this to say that the algorithm isn't suffering from some lack of floating-point precision because two vertices for every triangle in the mesh are shared between adjacent triangles. Therefore, the GPU's vertex cache should be kicking in and I shouldn't be seeing a decrease in FPS. </edit> [Edited by - hellraiser on September 18, 2007 6:32:32 PM]

Share this post


Link to post
Share on other sites
That's an insignificant difference in time.

865 fps = 1.15ms per frame

780 fps = 1.28ms per frame


that's a difference of 0.15ms (i.e. one one hundredth of one one thousandanth of a second.

i.e. there is effectively no difference in framerate.

-me

Share this post


Link to post
Share on other sites
Quote:
Original post by Palidine
That's an insignificant difference in time.

865 fps = 1.15ms per frame

780 fps = 1.28ms per frame


that's a difference of 0.15ms (i.e. one one hundredth of one one thousandanth of a second.

i.e. there is effectively no difference in framerate.

-me


Very much true but what's troubling me is the decrease in FPS in the first place. What happens when my scenes grow in complexity and likewise my graphics engine does too? Should I now maybe be thinking about changing algorithmic strategies and focus more on generating tri-strip meshes rather than triangles? Is this an isolated issue related to my graphics card alone?

I mean, I've got so many questions right now and no answers that it's making me doubt everything I've done so far in my modest graphics engine .

Thanks for your reply. :-)


PS: I've edited my original post and added some more info at the bottom.

Share this post


Link to post
Share on other sites
You're prematurely optimizing. There's no point in optimizing like this because it isn't a bottle-neck in your application. It currently makes absolutely no difference in the performance of your application whether you use strips or meshes. Therefore it doesn't matter (for now) what you choose.

Only when your application grows, and you detect through profiling that you need to revisit this decision is it a good time to optimize this part of your game.

-me

Share this post


Link to post
Share on other sites
Quote:
Original post by hellraiser
<edit>
The new skydome generating algorithm is in essence the same as it was before when it generated a tri-strip mesh. The only differences now is that the vertex buffer is larger so as to accomodate all the triangle vertices of the skydome ((vertices-2)*3) and also each vertice is stored at every 3rd position in the vertex buffer after the first 3 elements (vertexbuffer[n*3]=triangleVertice, n>3, 1 being the lowest index.) Then I iterate through the vertex buffer to finalize the triangles by using OpenGL's rules when rendering triangle_strips {odd=(n,n+1,n+2);even=(n+1,n,n+2)}.

All this to say that the algorithm isn't suffering from some lack of floating-point precision because two vertices for every triangle in the mesh are shared between adjacent triangles. Therefore, the GPU's vertex cache should be kicking in and I shouldn't be seeing a decrease in FPS.
</edit>


I'm slightly confused by this statement "The only differences now is that the vertex buffer is larger so as to accomodate all the triangle vertices of the skydome ((vertices-2)*3).

The vertex buffer shouldn't need to be any bigger, a triangle list in strip format uses the same amount of vertex data as a triangle strip does, the only difference is that it uses more index data.

So, for two shared triangles both method would have 4 vertices defined, however the triangle strip would have an index buffer of [0,1,2,3] and the triangle list would have an index buffer of [0,1,2,0,2,3].

The fact you don't mention an index buffer in any of your posts makes me doubt you are even using one; you should.
Simply setting positional information the same isn't enough to make use of the post-T&L cache; at data look up time, without an index, the GPU has no way of knowing that the data at position 4 is the same as the data at position 0. What the index list does is allow the GPU to say 'I know this data is the same, therefore I can use this stored result'.

I suspect you are rendering with glDrawArrays(), which is the slowest of the vertex array functions (well, the ones which don't pick the data one vertex at a time anyways), you should be using glDrawElements() or glDrawRangeElements(), these are MUCH faster due to the use of the index buffer (I don't have the results to hand right now, but I'm pretty sure in a vertex shader heavy scene I was seeing ~10x improvement between glDrawArrays and glDrawElements for the data submission).

In short;
- You need to use indices
- You don't need to generate more data

Share this post


Link to post
Share on other sites
Quote:
I'm slightly confused by this statement "The only differences now is that the vertex buffer is larger so as to accomodate all the triangle vertices of the skydome ((vertices-2)*3).

The vertex buffer shouldn't need to be any bigger, a triangle list in strip format uses the same amount of vertex data as a triangle strip does, the only difference is that it uses more index data.

So, for two shared triangles both method would have 4 vertices defined, however the triangle strip would have an index buffer of [0,1,2,3] and the triangle list would have an index buffer of [0,1,2,0,2,3].

The fact you don't mention an index buffer in any of your posts makes me doubt you are even using one; you should.


You are absolutely right; I'm not! :) That's why I've expanded the vertex buffer to store 3 vertices/triangle. I can now see what an idiot I was.

Quote:
Simply setting positional information the same isn't enough to make use of the post-T&L cache; at data look up time, without an index, the GPU has no way of knowing that the data at position 4 is the same as the data at position 0. What the index list does is allow the GPU to say 'I know this data is the same, therefore I can use this stored result'.

So that's how the vertex cache works...

In all honesty I always thought using index lists was an unnecessary waste of bandwidth, but then again I never quite understood the benefits from using them in the first place.

Quote:
I suspect you are rendering with glDrawArrays() [...]

Again, right on the money!

Quote:
[...] which is the slowest of the vertex array functions (well, the ones which don't pick the data one vertex at a time anyways), you should be using glDrawElements() or glDrawRangeElements(), these are MUCH faster due to the use of the index buffer (I don't have the results to hand right now, but I'm pretty sure in a vertex shader heavy scene I was seeing ~10x improvement between glDrawArrays and glDrawElements for the data submission).

In short;
- You need to use indices
- You don't need to generate more data

Thank you ever so much for the eye opener. There's not much I can say but to slap myself on the wrist...

You have no idea how helpful your post was to me, Phantom! Thanks again!

[Edited by - hellraiser on September 18, 2007 8:31:27 PM]

Share this post


Link to post
Share on other sites
Quote:
Original post by hellraiser
I've just converted a class that generated a triangle-strip mesh of a skydome to generating it with triangles. I did this as I've read a couple of articles that state that rendering triangles is slightly faster than strips because the GPU is able to take advantage of its fast vertex cache. Also, numerous posts here on GameDev by many gurus state just the same.
Who and where said that? In the last months/years, I never heard people saying there's a good speed increment.

Vertex cache is independant from the primitive. What's probably happening is that your tristrips are so long they trash your vcache.
Generating vcache aware strips for arbitrary geometry isn't trivial.

This is just to make clear that this tri-over-strip stuff is a myth.

Quote:
Original post by Palidine
You're prematurely optimizing.
Reasons: many algorithms doesn't scale linearly and graphics systems usually underperform at lower than average complexity.

Share this post


Link to post
Share on other sites
About that vertex cache...
I've just done some tests on line-strips vs lines. I expected drawing linestrips would be faster since I was inputting half the amount of vertices compared to drawing GL_LINES, but indead drawing 1 million lines took almost exactly the same time using GL_LINE_STRIP as GL_LINES.
So, my question I guess is, how does the vertex cache work? Why does it not help me in drawing ridiculously long LINE_STRIPS?

Erik Sintorn

Share this post


Link to post
Share on other sites
When it comes to drawing lines there is one very important point you need to remember: GPUs are truely rubbish at doing so.

This may come as a surprise, however if you consider it it's not that suprising; GPUs are optimised for the most common case, which is drawing triangles. While I can't recall the specifics off the top of my head drawing lines really causes issues, but as most games don't require them it really isn't a problem in the grand scheme of things.

Now, the vertex cache, this comes in two flavours;
1) pre-transform
2) post-transform

Flavour 1 just helps with data transfer, for the most part you'll never have to care about it.

Flavour 2 is the one we are intrested in. While it might be a little more complex than this in reality for practical purposes you can think of it as a kinda of map (an array of key-value pairs).

Once a vertex has been transformed by a vertex shader its output data is store in this array and the 'key' is set to the index. When the graphics card next goes to pull a piece of data for processing it will use the index of the vertex its about to deal with and first check if its in the cache. If it is then it reuses that data, if not then it fetches the data and performs the transform.

This array has a certain number of entries it can hold, so once it's full and it has another vertex to add it has to remove an entry; this is most probably done with a 'least-recently-used' scheme; so the vertex data which was accessed/stored longest ago is replaced with the new data.

So, if we take a VERY simple GPU which processes one vertex at a time and can cache 3 entries and apply it to my earlier example;
Quote:

So, for two shared triangles both method would have 4 vertices defined, however the triangle strip would have an index buffer of [0,1,2,3] and the triangle list would have an index buffer of [0,1,2,0,2,3].


It would go something like this (for the triangle list);
Quote:

Vertex 0 - not in cache; transform and store
Vertex 1 - not in cache; transform and store
Vertex 2 - not in cache; transform and store
Vertex 0 - in cache; reuse data
Vertex 2 - in cache; reuse data
Vertex 3 - not in cache; transform, cache full, remove entry 1, store

Note how the last vertex hits a full cache and dumps vertex 1 from it.

Now, in the hardware reality GPUs generally have more vertices in flight for processing than 1 at a time (my X1900XT for example can process 8 at once) and have larger caches (16 to 32 entries, this might well depend on the amount of outputs from a vertex shader), but the general priniple holds.

This is why you want to try to arrange your indices so you rehit as many vertices as possible to make use of the cache and reduce vertex processing overhead.

Of course, this does some what assume your bottleneck is in the vertex shader stage; if your pixel shader is doing so much heavy work that it dwarfs the vertex shader time then you'll want to focus your efforts there.

Share this post


Link to post
Share on other sites
EDIT: beaten by Phantom on time, which provides more detail. You can probably skip this message.

???

The vcache usually works in a temporal coherence driven way. It is usually a modified FIFO in which reused vertices are refreshed (that's LRU).

So... like all caches... if you draw 1M vertices like this:
--*--*--*--*--*--*--*--*--*--*--*
Your vcache is not going to give you any performance increment since there's nothing to cache in the first place (know your data!).
Also, if there's something to cache but you run off the FIFO the vcache won't help.

Said that, I noticed the mileage of using vcache-optimized geometry varies... It isn't different from other bottleneck analisys and bonks again on the concept that strips are not faster than tris (nor vice versa) by themselves.

In this specific case however, we should also recall that most consumer hardware is geared towards filled triangles and not lines so it wasn't probably a meaningful measurement in the first place.

Share this post


Link to post
Share on other sites
Quote:

In short;
- You need to use indices
- You don't need to generate more data


So what would you suggest if you have normals with everything? index it anyway? even though every vertex is going to be unique with the normal information?

I figured it was pretty useless so I decided to make everything interlaced arrays (GL_T2F_N3F_V3F) just to keep it consistent.


Share this post


Link to post
Share on other sites
Quote:
Original post by Jenison
Quote:

In short;
- You need to use indices
- You don't need to generate more data


So what would you suggest if you have normals with everything? index it anyway? even though every vertex is going to be unique with the normal information?


Ah, but not every vertex is going to have a unique normal. The majority of the normals will be built by say averaging the face normals of the faces the vertex is a part of (typical for terrain). A typical model will have very few vertices which require unique normals at the same position in space.

Put it another way; if indexing was no use for large scale models it wouldn't be used. Take a typical Doom3 model; zfat.md5mesh

This model has 874 vertices but using index data it has 1274 triangles. That's alot of shared vertices by anyones count, and these vertices all have normal information attached.

Remember a vertex ISNT just position; it is position, normal, colour, texture coords and anything else requred for the vertex transformation stage. Most of that informtion will be shared between triangles, in some cases it wont be but at that point you have to duplicate the other information because you no longer have a unique vertex.

Quote:

I figured it was pretty useless so I decided to make everything interlaced arrays (GL_T2F_N3F_V3F) just to keep it consistent.


The interlacing only effects how the gfx card expects the vertex data layed out in memory, it has no bearing on index use or not. (and without index data you are not using the post-T&L cache).

Share this post


Link to post
Share on other sites
Quote:
Phantom wrote:
<snip>
Once a vertex has been transformed by a vertex shader its output data is store in this array and the 'key' is set to the index. When the graphics card next goes to pull a piece of data for processing it will use the index of the vertex its about to deal with and first check if its in the cache. If it is then it reuses that data, if not then it fetches the data and performs the transform.
<snip>


Say you have a square grid of x by y tiles where each tile is painted with one texture, what would you say is the most efficient way to render it?

As I see it vertices can't be shared among adjacent tiles because the texture coordinates differ at adjacent vertices, which means there will have to be unique vertex data for every single tile in the mesh. Would you say using glMultiDrawElements is the most efficient way to render this grid? I think this would imply having 6 indices for the 4 vertices (2 tris) in each tile and then a consecutive list of indices for each tile.

BTW, brilliant explanation on how the vertex cache actually operates, Phantom. Thanks!

Share this post


Link to post
Share on other sites
Ah, that would be the 'worse case' situation for the cache.

glDrawElements would be all you need; Each group of two triangles makes a tile and as each tile in the current draw must come from the same bound texture you can build an index array for ALL the titles which use data from that texture to be drawn in one shot. (Batching by texture is a good thing, it saves unneeded state changes on the GPU and texture changes are the 3rd most expensive change, with pixel shader and vertex shader being 1 and 2 respectively).

The basic drawing would be:

for each texture to be draw
- bind texture
- draw vertex array with indices for that texture
next texture

If all the tiles are from the same texture then you can just setup the array and draw them all at once, batching by texture isn't an issue (although as a total sidenote I wonder if batching by tile in a texture makes much difference as it would be more texture cache friendly... hmmm).

Share this post


Link to post
Share on other sites
Thank you all for your answers, although I'm still a little confused so here goes. The reason I thought line strips would be faster than lines is that at least one vertex will ALLWAYS be in the cache? If I draw a line like:

*----*----*----*----*----*----*

The first vertex of each segment (save the first one) will already have been transformed by the segment before it when the GS needs it? By the way, I don't ever run the fragment shader so that's not what's bottlenecking me.

Share this post


Link to post
Share on other sites

This topic is 3737 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.

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

Sign in to follow this  

  • Forum Statistics

    • Total Topics
      628666
    • Total Posts
      2984131
  • Similar Content

    • By Vortez
      Hi guys, im having a little problem fixing a bug in my program since i multi-threaded it. The app is a little video converter i wrote for fun. To help you understand the problem, ill first explain how the program is made. Im using Delphi to do the GUI/Windows part of the code, then im loading a c++ dll for the video conversion. The problem is not related to the video conversion, but with OpenGL only. The code work like this:

       
      DWORD WINAPI JobThread(void *params) { for each files { ... _ConvertVideo(input_name, output_name); } } void EXP_FUNC _ConvertVideo(char *input_fname, char *output_fname) { // Note that im re-initializing and cleaning up OpenGL each time this function is called... CGLEngine GLEngine; ... // Initialize OpenGL GLEngine.Initialize(render_wnd); GLEngine.CreateTexture(dst_width, dst_height, 4); // decode the video and render the frames... for each frames { ... GLEngine.UpdateTexture(pY, pU, pV); GLEngine.Render(); } cleanup: GLEngine.DeleteTexture(); GLEngine.Shutdown(); // video cleanup code... }  
      With a single thread, everything work fine. The problem arise when im starting the thread for a second time, nothing get rendered, but the encoding work fine. For example, if i start the thread with 3 files to process, all of them render fine, but if i start the thread again (with the same batch of files or not...), OpenGL fail to render anything.
      Im pretty sure it has something to do with the rendering context (or maybe the window DC?). Here a snippet of my OpenGL class:
      bool CGLEngine::Initialize(HWND hWnd) { hDC = GetDC(hWnd); if(!SetupPixelFormatDescriptor(hDC)){ ReleaseDC(hWnd, hDC); return false; } hRC = wglCreateContext(hDC); wglMakeCurrent(hDC, hRC); // more code ... return true; } void CGLEngine::Shutdown() { // some code... if(hRC){wglDeleteContext(hRC);} if(hDC){ReleaseDC(hWnd, hDC);} hDC = hRC = NULL; }  
      The full source code is available here. The most relevant files are:
      -OpenGL class (header / source)
      -Main code (header / source)
       
      Thx in advance if anyone can help me.
    • By DiligentDev
      This article uses material originally posted on Diligent Graphics web site.
      Introduction
      Graphics APIs have come a long way from small set of basic commands allowing limited control of configurable stages of early 3D accelerators to very low-level programming interfaces exposing almost every aspect of the underlying graphics hardware. Next-generation APIs, Direct3D12 by Microsoft and Vulkan by Khronos are relatively new and have only started getting widespread adoption and support from hardware vendors, while Direct3D11 and OpenGL are still considered industry standard. New APIs can provide substantial performance and functional improvements, but may not be supported by older hardware. An application targeting wide range of platforms needs to support Direct3D11 and OpenGL. New APIs will not give any advantage when used with old paradigms. It is totally possible to add Direct3D12 support to an existing renderer by implementing Direct3D11 interface through Direct3D12, but this will give zero benefits. Instead, new approaches and rendering architectures that leverage flexibility provided by the next-generation APIs are expected to be developed.
      There are at least four APIs (Direct3D11, Direct3D12, OpenGL/GLES, Vulkan, plus Apple's Metal for iOS and osX platforms) that a cross-platform 3D application may need to support. Writing separate code paths for all APIs is clearly not an option for any real-world application and the need for a cross-platform graphics abstraction layer is evident. The following is the list of requirements that I believe such layer needs to satisfy:
      Lightweight abstractions: the API should be as close to the underlying native APIs as possible to allow an application leverage all available low-level functionality. In many cases this requirement is difficult to achieve because specific features exposed by different APIs may vary considerably. Low performance overhead: the abstraction layer needs to be efficient from performance point of view. If it introduces considerable amount of overhead, there is no point in using it. Convenience: the API needs to be convenient to use. It needs to assist developers in achieving their goals not limiting their control of the graphics hardware. Multithreading: ability to efficiently parallelize work is in the core of Direct3D12 and Vulkan and one of the main selling points of the new APIs. Support for multithreading in a cross-platform layer is a must. Extensibility: no matter how well the API is designed, it still introduces some level of abstraction. In some cases the most efficient way to implement certain functionality is to directly use native API. The abstraction layer needs to provide seamless interoperability with the underlying native APIs to provide a way for the app to add features that may be missing. Diligent Engine is designed to solve these problems. Its main goal is to take advantages of the next-generation APIs such as Direct3D12 and Vulkan, but at the same time provide support for older platforms via Direct3D11, OpenGL and OpenGLES. Diligent Engine exposes common C++ front-end for all supported platforms and provides interoperability with underlying native APIs. It also supports integration with Unity and is designed to be used as graphics subsystem in a standalone game engine, Unity native plugin or any other 3D application. Full source code is available for download at GitHub and is free to use.
      Overview
      Diligent Engine API takes some features from Direct3D11 and Direct3D12 as well as introduces new concepts to hide certain platform-specific details and make the system easy to use. It contains the following main components:
      Render device (IRenderDevice  interface) is responsible for creating all other objects (textures, buffers, shaders, pipeline states, etc.).
      Device context (IDeviceContext interface) is the main interface for recording rendering commands. Similar to Direct3D11, there are immediate context and deferred contexts (which in Direct3D11 implementation map directly to the corresponding context types). Immediate context combines command queue and command list recording functionality. It records commands and submits the command list for execution when it contains sufficient number of commands. Deferred contexts are designed to only record command lists that can be submitted for execution through the immediate context.
      An alternative way to design the API would be to expose command queue and command lists directly. This approach however does not map well to Direct3D11 and OpenGL. Besides, some functionality (such as dynamic descriptor allocation) can be much more efficiently implemented when it is known that a command list is recorded by a certain deferred context from some thread.
      The approach taken in the engine does not limit scalability as the application is expected to create one deferred context per thread, and internally every deferred context records a command list in lock-free fashion. At the same time this approach maps well to older APIs.
      In current implementation, only one immediate context that uses default graphics command queue is created. To support multiple GPUs or multiple command queue types (compute, copy, etc.), it is natural to have one immediate contexts per queue. Cross-context synchronization utilities will be necessary.
      Swap Chain (ISwapChain interface). Swap chain interface represents a chain of back buffers and is responsible for showing the final rendered image on the screen.
      Render device, device contexts and swap chain are created during the engine initialization.
      Resources (ITexture and IBuffer interfaces). There are two types of resources - textures and buffers. There are many different texture types (2D textures, 3D textures, texture array, cubmepas, etc.) that can all be represented by ITexture interface.
      Resources Views (ITextureView and IBufferView interfaces). While textures and buffers are mere data containers, texture views and buffer views describe how the data should be interpreted. For instance, a 2D texture can be used as a render target for rendering commands or as a shader resource.
      Pipeline State (IPipelineState interface). GPU pipeline contains many configurable stages (depth-stencil, rasterizer and blend states, different shader stage, etc.). Direct3D11 uses coarse-grain objects to set all stage parameters at once (for instance, a rasterizer object encompasses all rasterizer attributes), while OpenGL contains myriad functions to fine-grain control every individual attribute of every stage. Both methods do not map very well to modern graphics hardware that combines all states into one monolithic state under the hood. Direct3D12 directly exposes pipeline state object in the API, and Diligent Engine uses the same approach.
      Shader Resource Binding (IShaderResourceBinding interface). Shaders are programs that run on the GPU. Shaders may access various resources (textures and buffers), and setting correspondence between shader variables and actual resources is called resource binding. Resource binding implementation varies considerably between different API. Diligent Engine introduces a new object called shader resource binding that encompasses all resources needed by all shaders in a certain pipeline state.
      API Basics
      Creating Resources
      Device resources are created by the render device. The two main resource types are buffers, which represent linear memory, and textures, which use memory layouts optimized for fast filtering. Graphics APIs usually have a native object that represents linear buffer. Diligent Engine uses IBuffer interface as an abstraction for a native buffer. To create a buffer, one needs to populate BufferDesc structure and call IRenderDevice::CreateBuffer() method as in the following example:
      BufferDesc BuffDesc; BufferDesc.Name = "Uniform buffer"; BuffDesc.BindFlags = BIND_UNIFORM_BUFFER; BuffDesc.Usage = USAGE_DYNAMIC; BuffDesc.uiSizeInBytes = sizeof(ShaderConstants); BuffDesc.CPUAccessFlags = CPU_ACCESS_WRITE; m_pDevice->CreateBuffer( BuffDesc, BufferData(), &m_pConstantBuffer ); While there is usually just one buffer object, different APIs use very different approaches to represent textures. For instance, in Direct3D11, there are ID3D11Texture1D, ID3D11Texture2D, and ID3D11Texture3D objects. In OpenGL, there is individual object for every texture dimension (1D, 2D, 3D, Cube), which may be a texture array, which may also be multisampled (i.e. GL_TEXTURE_2D_MULTISAMPLE_ARRAY). As a result there are nine different GL texture types that Diligent Engine may create under the hood. In Direct3D12, there is only one resource interface. Diligent Engine hides all these details in ITexture interface. There is only one  IRenderDevice::CreateTexture() method that is capable of creating all texture types. Dimension, format, array size and all other parameters are specified by the members of the TextureDesc structure:
      TextureDesc TexDesc; TexDesc.Name = "My texture 2D"; TexDesc.Type = TEXTURE_TYPE_2D; TexDesc.Width = 1024; TexDesc.Height = 1024; TexDesc.Format = TEX_FORMAT_RGBA8_UNORM; TexDesc.Usage = USAGE_DEFAULT; TexDesc.BindFlags = BIND_SHADER_RESOURCE | BIND_RENDER_TARGET | BIND_UNORDERED_ACCESS; TexDesc.Name = "Sample 2D Texture"; m_pRenderDevice->CreateTexture( TexDesc, TextureData(), &m_pTestTex ); If native API supports multithreaded resource creation, textures and buffers can be created by multiple threads simultaneously.
      Interoperability with native API provides access to the native buffer/texture objects and also allows creating Diligent Engine objects from native handles. It allows applications seamlessly integrate native API-specific code with Diligent Engine.
      Next-generation APIs allow fine level-control over how resources are allocated. Diligent Engine does not currently expose this functionality, but it can be added by implementing IResourceAllocator interface that encapsulates specifics of resource allocation and providing this interface to CreateBuffer() or CreateTexture() methods. If null is provided, default allocator should be used.
      Initializing the Pipeline State
      As it was mentioned earlier, Diligent Engine follows next-gen APIs to configure the graphics/compute pipeline. One big Pipelines State Object (PSO) encompasses all required states (all shader stages, input layout description, depth stencil, rasterizer and blend state descriptions etc.). This approach maps directly to Direct3D12/Vulkan, but is also beneficial for older APIs as it eliminates pipeline misconfiguration errors. With many individual calls tweaking various GPU pipeline settings it is very easy to forget to set one of the states or assume the stage is already properly configured when in fact it is not. Using pipeline state object helps avoid these problems as all stages are configured at once.
      Creating Shaders
      While in earlier APIs shaders were bound separately, in the next-generation APIs as well as in Diligent Engine shaders are part of the pipeline state object. The biggest challenge when authoring shaders is that Direct3D and OpenGL/Vulkan use different shader languages (while Apple uses yet another language in their Metal API). Maintaining two versions of every shader is not an option for real applications and Diligent Engine implements shader source code converter that allows shaders authored in HLSL to be translated to GLSL. To create a shader, one needs to populate ShaderCreationAttribs structure. SourceLanguage member of this structure tells the system which language the shader is authored in:
      SHADER_SOURCE_LANGUAGE_DEFAULT - The shader source language matches the underlying graphics API: HLSL for Direct3D11/Direct3D12 mode, and GLSL for OpenGL and OpenGLES modes. SHADER_SOURCE_LANGUAGE_HLSL - The shader source is in HLSL. For OpenGL and OpenGLES modes, the source code will be converted to GLSL. SHADER_SOURCE_LANGUAGE_GLSL - The shader source is in GLSL. There is currently no GLSL to HLSL converter, so this value should only be used for OpenGL and OpenGLES modes. There are two ways to provide the shader source code. The first way is to use Source member. The second way is to provide a file path in FilePath member. Since the engine is entirely decoupled from the platform and the host file system is platform-dependent, the structure exposes pShaderSourceStreamFactory member that is intended to provide the engine access to the file system. If FilePath is provided, shader source factory must also be provided. If the shader source contains any #include directives, the source stream factory will also be used to load these files. The engine provides default implementation for every supported platform that should be sufficient in most cases. Custom implementation can be provided when needed.
      When sampling a texture in a shader, the texture sampler was traditionally specified as separate object that was bound to the pipeline at run time or set as part of the texture object itself. However, in most cases it is known beforehand what kind of sampler will be used in the shader. Next-generation APIs expose new type of sampler called static sampler that can be initialized directly in the pipeline state. Diligent Engine exposes this functionality: when creating a shader, textures can be assigned static samplers. If static sampler is assigned, it will always be used instead of the one initialized in the texture shader resource view. To initialize static samplers, prepare an array of StaticSamplerDesc structures and initialize StaticSamplers and NumStaticSamplers members. Static samplers are more efficient and it is highly recommended to use them whenever possible. On older APIs, static samplers are emulated via generic sampler objects.
      The following is an example of shader initialization:
      ShaderCreationAttribs Attrs; Attrs.Desc.Name = "MyPixelShader"; Attrs.FilePath = "MyShaderFile.fx"; Attrs.SearchDirectories = "shaders;shaders\\inc;"; Attrs.EntryPoint = "MyPixelShader"; Attrs.Desc.ShaderType = SHADER_TYPE_PIXEL; Attrs.SourceLanguage = SHADER_SOURCE_LANGUAGE_HLSL; BasicShaderSourceStreamFactory BasicSSSFactory(Attrs.SearchDirectories); Attrs.pShaderSourceStreamFactory = &BasicSSSFactory; ShaderVariableDesc ShaderVars[] = {     {"g_StaticTexture", SHADER_VARIABLE_TYPE_STATIC},     {"g_MutableTexture", SHADER_VARIABLE_TYPE_MUTABLE},     {"g_DynamicTexture", SHADER_VARIABLE_TYPE_DYNAMIC} }; Attrs.Desc.VariableDesc = ShaderVars; Attrs.Desc.NumVariables = _countof(ShaderVars); Attrs.Desc.DefaultVariableType = SHADER_VARIABLE_TYPE_STATIC; StaticSamplerDesc StaticSampler; StaticSampler.Desc.MinFilter = FILTER_TYPE_LINEAR; StaticSampler.Desc.MagFilter = FILTER_TYPE_LINEAR; StaticSampler.Desc.MipFilter = FILTER_TYPE_LINEAR; StaticSampler.TextureName = "g_MutableTexture"; Attrs.Desc.NumStaticSamplers = 1; Attrs.Desc.StaticSamplers = &StaticSampler; ShaderMacroHelper Macros; Macros.AddShaderMacro("USE_SHADOWS", 1); Macros.AddShaderMacro("NUM_SHADOW_SAMPLES", 4); Macros.Finalize(); Attrs.Macros = Macros; RefCntAutoPtr<IShader> pShader; m_pDevice->CreateShader( Attrs, &pShader );
      Creating the Pipeline State Object
      After all required shaders are created, the rest of the fields of the PipelineStateDesc structure provide depth-stencil, rasterizer, and blend state descriptions, the number and format of render targets, input layout format, etc. For instance, rasterizer state can be described as follows:
      PipelineStateDesc PSODesc; RasterizerStateDesc &RasterizerDesc = PSODesc.GraphicsPipeline.RasterizerDesc; RasterizerDesc.FillMode = FILL_MODE_SOLID; RasterizerDesc.CullMode = CULL_MODE_NONE; RasterizerDesc.FrontCounterClockwise = True; RasterizerDesc.ScissorEnable = True; RasterizerDesc.AntialiasedLineEnable = False; Depth-stencil and blend states are defined in a similar fashion.
      Another important thing that pipeline state object encompasses is the input layout description that defines how inputs to the vertex shader, which is the very first shader stage, should be read from the memory. Input layout may define several vertex streams that contain values of different formats and sizes:
      // Define input layout InputLayoutDesc &Layout = PSODesc.GraphicsPipeline.InputLayout; LayoutElement TextLayoutElems[] = {     LayoutElement( 0, 0, 3, VT_FLOAT32, False ),     LayoutElement( 1, 0, 4, VT_UINT8, True ),     LayoutElement( 2, 0, 2, VT_FLOAT32, False ), }; Layout.LayoutElements = TextLayoutElems; Layout.NumElements = _countof( TextLayoutElems ); Finally, pipeline state defines primitive topology type. When all required members are initialized, a pipeline state object can be created by IRenderDevice::CreatePipelineState() method:
      // Define shader and primitive topology PSODesc.GraphicsPipeline.PrimitiveTopologyType = PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE; PSODesc.GraphicsPipeline.pVS = pVertexShader; PSODesc.GraphicsPipeline.pPS = pPixelShader; PSODesc.Name = "My pipeline state"; m_pDev->CreatePipelineState(PSODesc, &m_pPSO); When PSO object is bound to the pipeline, the engine invokes all API-specific commands to set all states specified by the object. In case of Direct3D12 this maps directly to setting the D3D12 PSO object. In case of Direct3D11, this involves setting individual state objects (such as rasterizer and blend states), shaders, input layout etc. In case of OpenGL, this requires a number of fine-grain state tweaking calls. Diligent Engine keeps track of currently bound states and only calls functions to update these states that have actually changed.
      Binding Shader Resources
      Direct3D11 and OpenGL utilize fine-grain resource binding models, where an application binds individual buffers and textures to certain shader or program resource binding slots. Direct3D12 uses a very different approach, where resource descriptors are grouped into tables, and an application can bind all resources in the table at once by setting the table in the command list. Resource binding model in Diligent Engine is designed to leverage this new method. It introduces a new object called shader resource binding that encapsulates all resource bindings required for all shaders in a certain pipeline state. It also introduces the classification of shader variables based on the frequency of expected change that helps the engine group them into tables under the hood:
      Static variables (SHADER_VARIABLE_TYPE_STATIC) are variables that are expected to be set only once. They may not be changed once a resource is bound to the variable. Such variables are intended to hold global constants such as camera attributes or global light attributes constant buffers. Mutable variables (SHADER_VARIABLE_TYPE_MUTABLE) define resources that are expected to change on a per-material frequency. Examples may include diffuse textures, normal maps etc. Dynamic variables (SHADER_VARIABLE_TYPE_DYNAMIC) are expected to change frequently and randomly. Shader variable type must be specified during shader creation by populating an array of ShaderVariableDesc structures and initializing ShaderCreationAttribs::Desc::VariableDesc and ShaderCreationAttribs::Desc::NumVariables members (see example of shader creation above).
      Static variables cannot be changed once a resource is bound to the variable. They are bound directly to the shader object. For instance, a shadow map texture is not expected to change after it is created, so it can be bound directly to the shader:
      PixelShader->GetShaderVariable( "g_tex2DShadowMap" )->Set( pShadowMapSRV ); Mutable and dynamic variables are bound via a new Shader Resource Binding object (SRB) that is created by the pipeline state (IPipelineState::CreateShaderResourceBinding()):
      m_pPSO->CreateShaderResourceBinding(&m_pSRB); Note that an SRB is only compatible with the pipeline state it was created from. SRB object inherits all static bindings from shaders in the pipeline, but is not allowed to change them.
      Mutable resources can only be set once for every instance of a shader resource binding. Such resources are intended to define specific material properties. For instance, a diffuse texture for a specific material is not expected to change once the material is defined and can be set right after the SRB object has been created:
      m_pSRB->GetVariable(SHADER_TYPE_PIXEL, "tex2DDiffuse")->Set(pDiffuseTexSRV); In some cases it is necessary to bind a new resource to a variable every time a draw command is invoked. Such variables should be labeled as dynamic, which will allow setting them multiple times through the same SRB object:
      m_pSRB->GetVariable(SHADER_TYPE_VERTEX, "cbRandomAttribs")->Set(pRandomAttrsCB); Under the hood, the engine pre-allocates descriptor tables for static and mutable resources when an SRB objcet is created. Space for dynamic resources is dynamically allocated at run time. Static and mutable resources are thus more efficient and should be used whenever possible.
      As you can see, Diligent Engine does not expose low-level details of how resources are bound to shader variables. One reason for this is that these details are very different for various APIs. The other reason is that using low-level binding methods is extremely error-prone: it is very easy to forget to bind some resource, or bind incorrect resource such as bind a buffer to the variable that is in fact a texture, especially during shader development when everything changes fast. Diligent Engine instead relies on shader reflection system to automatically query the list of all shader variables. Grouping variables based on three types mentioned above allows the engine to create optimized layout and take heavy lifting of matching resources to API-specific resource location, register or descriptor in the table.
      This post gives more details about the resource binding model in Diligent Engine.
      Setting the Pipeline State and Committing Shader Resources
      Before any draw or compute command can be invoked, the pipeline state needs to be bound to the context:
      m_pContext->SetPipelineState(m_pPSO); Under the hood, the engine sets the internal PSO object in the command list or calls all the required native API functions to properly configure all pipeline stages.
      The next step is to bind all required shader resources to the GPU pipeline, which is accomplished by IDeviceContext::CommitShaderResources() method:
      m_pContext->CommitShaderResources(m_pSRB, COMMIT_SHADER_RESOURCES_FLAG_TRANSITION_RESOURCES); The method takes a pointer to the shader resource binding object and makes all resources the object holds available for the shaders. In the case of D3D12, this only requires setting appropriate descriptor tables in the command list. For older APIs, this typically requires setting all resources individually.
      Next-generation APIs require the application to track the state of every resource and explicitly inform the system about all state transitions. For instance, if a texture was used as render target before, while the next draw command is going to use it as shader resource, a transition barrier needs to be executed. Diligent Engine does the heavy lifting of state tracking.  When CommitShaderResources() method is called with COMMIT_SHADER_RESOURCES_FLAG_TRANSITION_RESOURCES flag, the engine commits and transitions resources to correct states at the same time. Note that transitioning resources does introduce some overhead. The engine tracks state of every resource and it will not issue the barrier if the state is already correct. But checking resource state is an overhead that can sometimes be avoided. The engine provides IDeviceContext::TransitionShaderResources() method that only transitions resources:
      m_pContext->TransitionShaderResources(m_pPSO, m_pSRB); In some scenarios it is more efficient to transition resources once and then only commit them.
      Invoking Draw Command
      The final step is to set states that are not part of the PSO, such as render targets, vertex and index buffers. Diligent Engine uses Direct3D11-syle API that is translated to other native API calls under the hood:
      ITextureView *pRTVs[] = {m_pRTV}; m_pContext->SetRenderTargets(_countof( pRTVs ), pRTVs, m_pDSV); // Clear render target and depth buffer const float zero[4] = {0, 0, 0, 0}; m_pContext->ClearRenderTarget(nullptr, zero); m_pContext->ClearDepthStencil(nullptr, CLEAR_DEPTH_FLAG, 1.f); // Set vertex and index buffers IBuffer *buffer[] = {m_pVertexBuffer}; Uint32 offsets[] = {0}; Uint32 strides[] = {sizeof(MyVertex)}; m_pContext->SetVertexBuffers(0, 1, buffer, strides, offsets, SET_VERTEX_BUFFERS_FLAG_RESET); m_pContext->SetIndexBuffer(m_pIndexBuffer, 0); Different native APIs use various set of function to execute draw commands depending on command details (if the command is indexed, instanced or both, what offsets in the source buffers are used etc.). For instance, there are 5 draw commands in Direct3D11 and more than 9 commands in OpenGL with something like glDrawElementsInstancedBaseVertexBaseInstance not uncommon. Diligent Engine hides all details with single IDeviceContext::Draw() method that takes takes DrawAttribs structure as an argument. The structure members define all attributes required to perform the command (primitive topology, number of vertices or indices, if draw call is indexed or not, if draw call is instanced or not, if draw call is indirect or not, etc.). For example:
      DrawAttribs attrs; attrs.IsIndexed = true; attrs.IndexType = VT_UINT16; attrs.NumIndices = 36; attrs.Topology = PRIMITIVE_TOPOLOGY_TRIANGLE_LIST; pContext->Draw(attrs); For compute commands, there is IDeviceContext::DispatchCompute() method that takes DispatchComputeAttribs structure that defines compute grid dimension.
      Source Code
      Full engine source code is available on GitHub and is free to use. The repository contains two samples, asteroids performance benchmark and example Unity project that uses Diligent Engine in native plugin.
      AntTweakBar sample is Diligent Engine’s “Hello World” example.

       
      Atmospheric scattering sample is a more advanced example. It demonstrates how Diligent Engine can be used to implement various rendering tasks: loading textures from files, using complex shaders, rendering to multiple render targets, using compute shaders and unordered access views, etc.

      Asteroids performance benchmark is based on this demo developed by Intel. It renders 50,000 unique textured asteroids and allows comparing performance of Direct3D11 and Direct3D12 implementations. Every asteroid is a combination of one of 1000 unique meshes and one of 10 unique textures.

      Finally, there is an example project that shows how Diligent Engine can be integrated with Unity.

      Future Work
      The engine is under active development. It currently supports Windows desktop, Universal Windows and Android platforms. Direct3D11, Direct3D12, OpenGL/GLES backends are now feature complete. Vulkan backend is coming next, and support for more platforms is planned.
    • By michaeldodis
      I've started building a small library, that can render pie menu GUI in legacy opengl, planning to add some traditional elements of course.
      It's interface is similar to something you'd see in IMGUI. It's written in C.
      Early version of the library
      I'd really love to hear anyone's thoughts on this, any suggestions on what features you'd want to see in a library like this? 
      Thanks in advance!
    • By Michael Aganier
      I have this 2D game which currently eats up to 200k draw calls per frame. The performance is acceptable, but I want a lot more than that. I need to batch my sprite drawing, but I'm not sure what's the best way in OpenGL 3.3 (to keep compatibility with older machines).
      Each individual sprite move independently almost every frame and their is a variety of textures and animations. What's the fastest way to render a lot of dynamic sprites? Should I map all my data to the GPU and update it all the time? Should I setup my data in the RAM and send it to the GPU all at once? Should I use one draw call per sprite and let the matrices apply the transformations or should I compute the transformations in a world vbo on the CPU so that they can be rendered by a single draw call?
    • By zolgoz
      Hi!

      I've recently started with opengl and just managed to write my first shader using the phong model. I want to learn more but I don't know where to begin, does anyone know of good articles or algorithms to start with?
      Thanks in advance.
  • Popular Now