Recent content by 991060

For anyone wants to implement megatexture himself, here's some useful information: http://www.cs.virginia.edu/~gfx/Courses/2002/BigData/papers/Texturing/Clipmap.pdf I'm not sure if John Carmack got his idea from this paper, but it is very similar to what megatexture provides.

How about finding the transformation matrix through vertices before and after transformation, then isolating translation/rotation/scaling/cropping components?

unrolling loops will lead to instruction count explosion, if the struct within loop is very complex or the loop count is high, hence the driver may not always do unrolling. There's no need to replace static loop with dynamic loop, because it gains you nothing. I assume X1K execute static loop...

zeckensack, you're right about the usage of filtering on vertex texturing. I simply overlook the situation in which the resolution of vertex and vertex texture is not 1:1. Yet it raises another question: how to compute the correct LOD on for arbitrary meshes? This kind of computation requires...

The texture filtering algorithm depends on the size and position of the fragments on the screen. When rendering to VB, you're effectively rendering to a Nx1 render target, I don't know how the filtering unit can do any reasonable work in this situation. Also, a vertex is a mathmatical...

A few questions concerning R2VB: 1.how do I take advantage of texture filtering in PS unit when the topology information is lost during the R2VB pass? 2.Only in certain situations, R2VB can be faster than VTF: the texture can be directly used as a vertex buffer. In most other cases, when...

This is probably the most remarkable achievement of R5xx lines: http://techreport.com/reviews/2005q4/radeon-x1000/shadermark-shadow.gif

There're probably 2 reasons: 1. Carmack doesn't trust the Cg/GLSL compiler. I believe there's no driver scored 100% in the GLSL parse test ATM. And Cg doesn't compile quite good on ATI HW. 2. He wants total control on the instruction count. When the shaders get too complex, you'll find...

There's no need to use high level language to write simple shaders as those found in DOOM3.

Provided the constant register file is large enough to hold dozens of pose datas, you can provide each instance a different offset to the constant register in the per instance data, that effectively enables you loop through the constant as you draw each instance of the character.

How do I make sure managed DX9.0c is installed? I have 2005 June SDK, .net framework 2.0 beta installed, and it throws a file not found exception.

I think the vec unit is capable of 8 ops/clock(2 ops per channel), while the scalar unit can only do 1 ops/clock.

1.How do they recommend using the CPU L2 cache => GPU data path? 2.What does the "64 active threads" mean exactly? Isn't it not enough to cover the texture read latency? And how many TMUs does the GPU have? 3.What made them think the 4D+1D design is better than other alternative(such as a...

dispatch threads from a command queue, like the arbiter in the Xenon GPU patent?

can anyone get a clearer version of this slide? http://images.anandtech.com/reviews/video/Playstation3/E3PressConference/rsxarchitecture.jpg