NVidia FX1000

I've just ordered (and Dell shippped) a new 650 with an NVidia FX1000 card... Then on yesterday's Knowledge Base Monitor I saw a TAN that
the FX1000 has very poor performance when using Wildfire and transparent parts, and that the config.pro setting should be put back the the 'old' settings, where transparent parts looked more like screens than smooth.
Has anyone any firsthand experience with this? The SPR had a fixed datecode of 2003330, which makes this even more puzzling. I just junked a Wildcat 7110 because of issues with display corruption, so I've gone down a step to the FX1000, and now where am I, back to a Quadro4 900?
Regards Peter Brown Jarvis Products Corp
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
: I've just ordered (and Dell shippped) a new 650 with an NVidia FX1000 : card... Then on yesterday's Knowledge Base Monitor I saw a TAN that : the FX1000 has very poor performance when using Wildfire and : transparent parts, and that the config.pro setting should be put back : the the 'old' settings, where transparent parts looked more like : screens than smooth. : That's very strange. The FX1000 is a good workstation card from a good gpu maker. It was an upgrade of the Quadro 4 and is supposed to outperform the 900 by 30%, so going backwards doesn't seem to be an option. It is also on PTC's website as a supported card for the Dell 650 on both XP and Win2k platforms. To help deepen this mystery, let me tell you about the nVidia GeForce2 Go card on my Toshiba laptop. It does Wildfire graphics with no problems, including a sweet job with the transparency and the PhotoLux rendering. The card has obvious limitations with only 16megs of memory, but the basic functioning is good.
I'd sure like to know how they come up with the TANs after they approved a card which seems like it ought to be able to do everything but backflips. Is it based on one computer and one installation? Is some graphics testing company testing cards with Pro/e and how systematically? Is there a published set of graphics functions that a card/driver should have to perform well with Pro/e? I sure haven't seen any such thing from PTC, although Dell has published a list of OpenGL functions the nVidia card will perform.
So, what I'm saying is that I wouldn't panic and start buying other cards just yet. PTC might simply be playing cya with a big client. It's support style generally reminds me of the computer business's idea of support from the 80s where you'd call up Dell or Compaq and they'd say, no sorry, that's not our problem, you've got to call Microsoft; then you'd call MS and they'd say, no sorry, that's not our problem, you've got to talk to the peripheral vendor and get their latest driver, etc, etc, where 'support' is a finger pointing run-around. The answer to this has always been ~ hold their feet to the fire, make them give you what you paid for.
David Janes
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Well, I don't see what all the fuss was about. Wildfire transparency appears to work fine with only the expected slight degredation in performance because of the increased intensiveness of the application versus the old stippled method.
Hardware: Dell 650, 3.06 GHz (hyper threading off) 2Gb ram NVidia FX1000 UltraSharp 20" flat panel at 1600x1200 Wildfire 2003330
Tested it in a 75 component assembly.
Note to Vericut users: I am having lots of trouble on this platform with machine simulation view and Vericut exiting with a javaw.exe error. Different dispaly drivers and versions of the Java Runtime Environment installed, with no fix yet. Files are at PTC and are probably going to be forwarded to CGTech.
Regards Peter Brown Jarvis Products Corp
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.