<br><br><div class="gmail_quote">2008/11/21 Franz Marini <span dir="ltr"><<a href="mailto:franz.marini@mi.infn.it">franz.marini@mi.infn.it</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;">
H<br>
Regarding the proprietary-ness of CUDA, I would argue that being<br>
proprietary also means that it probably better targets the NV GPU<br>
architecture, and a more general, portable solution, like OpenCL (which<br>
seems to be closer than expected, by the way) will possibly mean a<br>
somewhat less optimal use of the GPU. M</blockquote><div><br>Guys, I'm going to be controversial here.<br>The market may SAY otherwise, but the market does not give a rat's behind about proprietariness.<br>Tell a scientist that her N-body dynamics astrophysics model will run 500 times faster on a certain GPU and<br>
she'll get more papers published and an invite to a conference in Hawaii next year and you'll see those<br>grant dollars being spent.<br>Tell and engineer that his Nastran model or his CFD simulation will finish whilst he goes off to lunch/coffee and he'll<br>
bite your hand off.<br><br>It all comes down to codes - when the ISV codes use these things, you'll see the uptake.<br><br><br></div></div><br>