NV GF6系列架构分析(GPU gem2 读书笔记)
http://www.cnblogs.com/wangdaniu/archive/2006/02/20/334089.html
总线带宽:PCI Express×8, 66Mhz *32bits(4Bytes)*8 = 4 GB/s
目前计算机中的带宽:
GPU显存:550Mhz DDR ×256bits(32Bytes)×2=35GB/s
PCI Express ×16 :8GB/s
CPU内存,800Mhz前端总线:6.4GB/s
显卡在PC体系中的位置:
CPU
/\
|
6.4GB/s
|
\/
内存<=6.4GB/s or more=>北桥<=up to 8GB/s=>GPU=> to display
/\ /\
| |
| up to 35GB/s
\/ |
南桥 \/
/\ 显存
|
\/
其他设备
图形操作流程:
Vertex Data => Vertex Shader×6 => Cull/Clip/Setup => Raster =>Pixel Shader ×16 =>Fragment Crossbar => Z/Blend => Back Buffer
Color Buffer => Pixel Shader (as texture)
Z Buffer=> Early Z Cull
Pixel Shader => Vertex Shader (maybe stream out)
顶点处理单元:
顶点数据==> 32位浮点向量ALU×1
32位浮点数值ALU×1==>分支单元==>返回ALU
顶点纹理采样单元×1 ==>或者生成Primitive==>culling/clipping==>光栅化
像素处理单元:
图元数据==>32位浮点ALU×1:执行一次乘加操作或者纹理采样==>
==>32位浮点ALU×1(一次加操作)==>分支单元==>返回ALU
==>或者进入雾化ALU