D3D9 GPU Hacks

I’ve been trying to catch up what hacks GPU vendors have exposed in Direct3D9, and turns out there’s a lot of them!

If you know more hacks or more details, please let me know in the comments!

Most hacks are exposed as custom (“FOURCC”) formats. So to check for that, you do CheckDeviceFormat. Here’s the list (Usage column codes: DS=DepthStencil, RT=RenderTarget; Resource column codes: tex=texture, surf=surface). More green = more hardware support.

Format Usage Resource Description NVIDIA GeForce ATI Radeon Intel
Shadow mapping
D3DFMT_D16 DS tex Sample depth buffer directly as shadow map. 3+ HD 2xxx+ 965+
D3DFMT_D24X8 DS tex 3+ HD 2xxx+ 965+
Depth Buffer As Texture
DF16 DS tex Read depth buffer as texture.   9500+ G45+
DF24 DS tex   X1300+ SB+
INTZ DS tex 8+ HD 4xxx+ G45+
RAWZ DS tex 6 & 7    
Anti-Aliasing related
RESZ RT surf Resolve MSAA’d depth stencil surface into non-MSAA’d depth texture.   HD 4xxx+ G45+
ATOC 0 surf Transparency anti-aliasing. 7+   SB+
SSAA 0 surf 7+    
All ATI SM2.0+ hardware   9500+  
n/a     Coverage Sampled Anti-Aliasing[6] 8+    
Texturing
ATI1 0 tex ATI1n & ATI2n texture compression formats. 8+ X1300+ G45+
ATI2 0 tex 6+ 9500+ G45+
DF24 DS tex Fetch 4: when sampling 1 channel texture, return four touched texel values[1]. Check for DF24 support.   X1300+ SB+
Misc
NULL RT surf Dummy render target surface that does not consume video memory. 6+ HD 4xxx+ HD+
NVDB 0 surf Depth Bounds Test. 6+    
R2VB 0 surf Render into vertex buffer. 6 & 7 9500+  
INST 0 surf Geometry Instancing on pre-SM3.0 hardware.   9500+  

Native Shadow Mapping

Native support for shadow map sampling & filtering was introduced ages ago (GeForce 3) by NVIDIA. Turns out ATI also implemented the same feature for it’s DX10 level cards. Intel also supports it on Intel 965 (aka GMA X3100, the shader model 3 card) and later (G45/X4500/HD) cards.

The usage is quite simple; just create a texture with regular depth/stencil format and render into it. When reading from the texture, one extra component in texture coordinates will be the depth to compare with. Compared & filtered result will be returned.

Also useful:

  • Creating NULL color surface to keep D3D runtime happy and save on video memory.

Depth Buffer as Texture

For some rendering schemes (anything with “deferred”) or some effects (SSAO, depth of field, volumetric fog, …) having access to a depth buffer is needed. If native depth buffer can be read as a texture, this saves both memory and a rendering pass or extra output for MRTs.

Depending on hardware, this can be achieved via INTZ, RAWZ, DF16 or DF24 formats:

  • INTZ is for recent (DX10+) hardware. With recent drivers, all three major IHVs expose this. According to ATI [1], it also allows using stencil buffer while rendering. Also allows reading from depth texture while it’s still being used for depth testing (but not depth writing). Looks like this applies to NV & Intel parts as well.
  • RAWZ is for GeForce 6 & 7 series only. Depth is specially encoded into four channels of returned value.
  • DF16 and DF24 is for ATI and Intel cards, including older cards that don’t support INTZ. Unlike INTZ, this does not allow using depth buffer or using the surface for both sampling & depth testing at the same time.

Also useful when using depth textures:

  • Creating NULL color surface to keep D3D runtime happy and save on video memory.
  • RESZ allows resolving multisampled depth surfaces into non-multisampled depth textures (result will be sample zero for each pixel).

Caveats:

  • Using INTZ for both depth/stencil testing and sampling at the same time seems to have performance problems on ATI cards (checked Radeon HD 3xxx to 5xxx, Catalyst 9.10 to 10.5). A workaround is to render to INTZ depth/stencil first, then use RESZ to “blit” it into another surface. Then do sampling from one surface, and depth testing on another.

Depth Bounds Test

Direct equivalent of GL_EXT_depth_bounds_test OpenGL extension. See [3] for more information.

Transparency Anti-Aliasing

NVIDIA exposes two controls: transparency multisampling (ATOC) and transparency supersampling (SSAA) [5]. ATI says that all Radeons since 9500 support “alpha to coverage” [1]. Intel supports ATOC with SandyBridge (GMA HD 2000/3000) GPUs.

Render Into Vertex Buffer

Similar to “stream out” or “memexport” in other APIs/platforms. See [2] for more information. Apparently some NVIDIA GPUs (or drivers?) support this as well.

Geometry Instancing

Instancing is supported on all Shader Model 3.0 hardware by Direct3D 9.0c, so there’s no extra hacks necessary there. ATI has exposed a capability to enable instancing on their Shader Model 2.0 hardware as well. Check for “INST” support, and do dev->SetRenderState (D3DRS_POINTSIZE, kFourccINST); at startup to enable instancing.

I can’t find any document on instancing from AMD now. Other references: [7] and [8].

ATI1n & ATI2n Compressed Texture Formats

Compressed texture formats. ATI1n is known as BC4 format in DirectX 10 land; ATI2n as BC5 or 3Dc. Since they are just DX10 formats, support for this is quite widespread, with NVIDIA exposing it a while ago and Intel exposing it recently (drivers 15.17 or higher).

Thing to keep in mind: when DX9 allocates the mip chain, they check if the format is a known compressed format and allocate the appropriate space for the smallest mip levels. For example, a 1x1 DXT1 compressed level actually takes up 8 bytes, as the block size is fixed at 4x4 texels. This is true for all block compressed formats. Now when using the hacked formats DX9 doesn’t know it’s a block compression format and will only allocate the number of bytes the mip would have taken, if it weren’t compressed. For example a 1x1 ATI1n format will only have 1 byte allocated. What you need to do is to stop the mip chain before the size of the either dimension shrinks below the block dimensions otherwise you risk having memory corruption.

Another thing to keep in mind: on Vista+ (WDDM) driver model, textures in these formats will still consume application address space. Most regular textures like DXT5 don’t take up additional address space in WDDM (see here). For some reason ATI1n and ATI2n textures on D3D9 are deemed lockable.

References

All this information gathered mostly from:

  1. Advanced DX9 Capabilities for ATI Radeon Cards (pdf)
  2. ATI R2VB Programming (pdf)
  3. NVIDIA GPU Programming Guide (pdf)
  4. ATI Tesselation
  5. NVIDIA Transparency AA
  6. NVIDIA Coverage Sampled AA
  7. Humus’ Instancing Demo
  8. Arseny’s article on particles

Changelog

  • 2013 06 11: One more note on ATI1n/ATI2n format virtual address space issue (thanks JSeb!).
  • 2013 04 09: Turns out since sometime 2011 Intel has DF24 and Fetch4 for SandyBridge and later.
  • 2011 01 09: Intel implemented ATOC for SandyBridge, and NULL for GMA HD and later.
  • 2010 08 25: Intel implemented DF16, INTZ, RESZ for G45+ GPUs!
  • 2010 08 25: Added note on INTZ performance issue with ATI cards.
  • 2010 08 19: Intel implemented ATI1n/ATI2n support for G45+ GPUs in the latest drivers!
  • 2010 07 08: Added note on ATI1n/ATI2n texture formats, with a caveat pointed out by Henning Semler (thanks!)
  • 2010 01 06: Hey, shadow map hacks are also supported on Intel 965!
  • 2009 12 09: Shadow map hacks are supported on Intel G45!
  • 2009 11 21: Added instancing on SM2.0 hardware.
  • 2009 11 20: Added Fetch-4, CSAA.
  • 2009 11 20: Initial version.

原文链接:http://aras-p.info/texts/D3D9GPUHacks.html

D3D9 GPU Hacks (转载)的更多相关文章

  1. D3D9 优化小技巧

    此篇文章主要讲一些小技巧,针对前面转载的D3D9 GPU Hacks,我们可以做的一些优化. 在做延迟渲染或者其它需要深度的地方使用INTZ格式的纹理,这样可以直接对纹理进行操作,节省了显存和带宽,这 ...

  2. 【转载】 GPU状态监测 nvidia-smi 命令详解

    原文地址: https://blog.csdn.net/huangfei711/article/details/79230446 ----------------------------------- ...

  3. 【转载】 os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID" os.environ["CUDA_VISIBLE_DEVICES"] = "0" (---------tensorflow中设置GPU可见顺序和选取)

    原文地址: https://blog.csdn.net/Jamesjjjjj/article/details/83414680 ------------------------------------ ...

  4. 【转载】GPU深度发掘(一)::GPGPU数学基础教程

    作者:Dominik Göddeke                 译者:华文广 Contents 介绍 准备条件 硬件设备要求 软件设备要求 两者选择 初始化OpenGL GLUT OpenGL ...

  5. 【转载】 NVIDIA Tesla/Quadro和GeForce GPU比较

    原文地址: https://blog.csdn.net/m0_37462765/article/details/74394932 版权声明:本文为博主原创文章,遵循 CC 4.0 BY-SA 版权协议 ...

  6. 【转载】GPU 加速下的图像处理

    Instagram,Snapchat,Photoshop. 所有这些应用都是用来做图像处理的.图像处理可以简单到把一张照片转换为灰度图,也可以复杂到是分析一个视频,并在人群中找到某个特定的人.尽管这些 ...

  7. 【Todo】【转载】深度学习&神经网络 科普及八卦 学习笔记 & GPU & SIMD

    上一篇文章提到了数据挖掘.机器学习.深度学习的区别:http://www.cnblogs.com/charlesblc/p/6159355.html 深度学习具体的内容可以看这里: 参考了这篇文章:h ...

  8. [转载]tensorflow中使用tf.ConfigProto()配置Session运行参数&&GPU设备指定

    tf.ConfigProto()函数用在创建session的时候,用来对session进行参数配置: config = tf.ConfigProto(allow_soft_placement=True ...

  9. 为什么GPU可以用于科学计算【转载】

    转自:https://blog.csdn.net/xihuanyuye/article/details/81178352 https://www.zhihu.com/question/35063258 ...

随机推荐

  1. 浏览器User-agent String里的历史故事

    你是否好奇标识浏览器身份的User-Agent,为什么每个浏览器都有Mozilla字样? Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 ...

  2. QT学习之-HelloWorld

    实现HelloWorld有2种方法. 第一种在mainwindow.ui直接添加Label且写上“HelloWorld”. 第二种是代码直接实现 #include "mainwindow.h ...

  3. S5PV210裸板驱动:启动

    以往2440和6410的启动方式,只要我们把裸板代码烧写到NAND FLASH的开始位置,当开发板上点启动时,处理器会自动从NAND FLASH上拷贝前面一段的代码到内部的RAM中执行.按照以前的方法 ...

  4. HTML: xhtml和html的區別

    XHTML 与 HTML 4.01 标准没有太多的不同. 另外,你应该马上使用小写字母编写HTML代码,同时绝不要养成忽略类似 </p> 标签的坏习惯. 最主要的不同: XHTML 元素必 ...

  5. java 客户端获取真实ip地址

    在开发工作中,我们常常需要获取客户端的IP.一般获取客户端的IP地址的方法是:request.getRemoteAddr();但是在通过了Apache,Squid等反向代理软件就不能获取到客户端的真实 ...

  6. php数据缓存

    用php进行微信开发时,碰到access_token长久保存的问题,以前都是用框架里的Cache直接set.get一下就完了.现在没框架可用了,只好自己动手写一个cache暂时用. 这个Cache类用 ...

  7. 【转】javascript中this的四种用法

    在函数执行时,this 总是指向调用该函数的对象.要判断 this 的指向,其实就是判断 this 所在的函数属于谁. 在<javaScript语言精粹>这本书中,把 this 出现的场景 ...

  8. delphi android 录像(调用Java的功能)

    delphi xe系列自带的控件都无法保存录像,经网友帮忙,昨天终于实现了录像功能(但有个问题是录像时无画面显示),程序主要使用了JMediaRecorder,MediaRecorder的使用方法可参 ...

  9. MongoDB过过瘾

    MongoDB 中默认的数据库为 test,连接后尝试以下操作 连接 插入数据:用过json的同学看到这格式相信不会陌生吧! db.person.insert({}) db.person.insert ...

  10. JS-006-表格元素操作

    直接上菜咯... 以下为 HTML 表格源码: <html> <head> <meta http-equiv="Content-Type" conte ...