| TIP: Click on subject to list as thread! | ANSI |
| echo: | |
|---|---|
| to: | |
| from: | |
| date: | |
| subject: | Firebird browser |
Hi Jasen! :-) JB> still it takes at best two write cycles to write a 24 bit pixel JB> (an 8 bit write and a 16 bit write) if the 16 bit write straddles a JB> data-bus boundary it'll take 3 cycles, and these cycles happen at the JB> video-bus speed not at the CPU speed. Well, with write gathering and what not between the CPU and bus, it is hard to say which writes occur in what ordering. If another write to the next pixel comes next, those four writes might be combined into one (or two) before sent over the bus. On the other hand, one could also first read full 32 bits, modify 24 bits and then write 32 bits back to the graphics card again. JB> with a 32 bit pixel it takes a single write and you're guaranteed to JB> always fit inside the bus bouandary. Likely, yes. Another consideration is framebuffer size... 32 bits means 1 byte per pixel is lost, which makes for smaller texture buffers and other things. JB> OTOH these days if you're pushing single pixels around using the CPU JB> you're probably doing whatever graphics task the slow way. That's for sure. :) JB> yeah I guess they could go for a 24 bit data bus on the video card, JB> or something like that... Well, current stuff usually has 256 bit or more internal data bus width. Ciao Pascal --- Msged/LNX 6.1.1* Origin: Freedom defined is freedom denied. (1:153/401.2) SEEN-BY: 633/267 270 @PATH: 153/401 307 140/1 106/2000 633/267 |
|
| SOURCE: echomail via fidonet.ozzmosis.com | |
Email questions or comments to sysop@ipingthereforeiam.com
All parts of this website painstakingly hand-crafted in the U.S.A.!
IPTIA BBS/MUD/Terminal/Game Server List, © 2025 IPTIA Consulting™.