<div dir="ltr"><br><br><div class="gmail_quote">2008/10/4 Ellis Wilson <span dir="ltr"><<a href="mailto:xclski@yahoo.com">xclski@yahoo.com</a>></span><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="Ih2E3d">Bruno Coutinho wrote:<br>
> You need to get data from machine main memory, compress and send results<br>
> back several times.<br>
> The bandwidth of pci express today is 8GB/s, so this is the maximum data<br>
> rate a gpu can compress.<br>
> You can use some tricks like computation and i/o (to main memory)<br>
> parallelization, but will be constrained to 8GB/s anyway.<br>
<br>
</div>Check out PCI Express 2 then, you'll get 16GB/s bandwidth but I'm not<br>
sure all the GPU's are moved that direction yet.<br>
<font color="#888888"></font></blockquote><div><br>This is in a x32 slot (that I never saw) in a typical x16 slot pci-e 1.0 make 4GB/s and 2.0 makes 8GB/s.<br> <br>All nvidia cards from 9 series and ATI from 3xxx series have pci express 2.0. <br>
<br>In chipsets only the newest ones like intel P45 have pci-e 2.0.<br><br></div></div><br></div>