comp.lang.idl-pvwave archive: archive

Home » Public Forums » archive » video cards

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

video cards [message #35142]

Tue, 20 May 2003 07:56

David.Chevrier
Messages: 14
Registered: May 2003

Junior Member

Hi - I was wondering if anyone is using a 3D Labs Wildcat4 7210 video
card. This is a high-end workstation that supports OpenGL, but not
DirectX (which is fine for me as this is my work computer and only
used for IDL). I wanted to know what people thought of this card and
how they set their video options. I'm not setting the results I had
hoped for... I know that some of this is due to both IDL memory
limitations on 32-bit OS's (I'm using WinXP pro).
http://www.rsinc.com/services/techtip.asp?ttid=3346 I would also like
to see what other video cards people are using and which they think
are the best. Any advice or input is appreciated. The rest of my rig
is: dual liquid-cooled (swiftech) 2.8 533fsb xeons, 4gb pc2100 ecc
reg. ddr, vantec stealth 520w power supply, 52x24x52 lite-on burner,
120gb 8mb cache western digital hd, supermicro x5dae mb, lian-li
pc-75b case with clear window. (This is my dream machine.... I did it
up with red led's and neon lights... complete with temperature monitor
and fan controller!) I write mostly object-oriented gui's for
government acoustic data at NOAA. Thanks!

Report message to a moderator

Re: video cards [message #35221 is a reply to message #35142]

Tue, 27 May 2003 09:46

Rick Towler
Messages: 821
Registered: August 1998

Senior Member

"David Chevrier" wrote in message...

> thanks for all the info!
>
>> Performance depends a lot on what you are doing. Can you describe
>> what you are trying to do?

> i'm working with HUGE arrays. we use a simrad ek500 echosounder

Just wait until you get that SM2000!

I can tell you that your setup is about as good as it is going to get. With
data this dense you will need to get creative because as you have discovered
you really can't just throw the data at the hardware and get the results you
want.

> rick - do you know Mike Jech?

Are you working with Mike "Groundskeeper Willie" Jech? He didn't buy you
that computer, did he? Looks like I need to upgrade...

Report message to a moderator

Re: video cards [message #35224 is a reply to message #35142]

Tue, 27 May 2003 07:43

David.Chevrier
Messages: 14
Registered: May 2003

Junior Member

thanks for all the info!

> Performance depends a lot on what you are doing. Can you describe what you
> are trying to do? How are you measuring performance?
>
> What sort of acoustics?

i'm working with HUGE arrays. we use a simrad ek500 echosounder
with 3 seperate frequencies, so i am getting 3 650 floating-point
arrays per second. i go out to sea and collect data 6 weeks a year,
so i end up with gb's and gb's of data. i was able to make a program
that allows you to interactly view weeks of data at once (using rebin
and congrid). the tough part was making a surface that had the x-y-z
axis as lng,lat,depth, where the boat traveled (with cutoffs at the
ocean depth) and then pasting the acoustic data onto the surface. (i
know this sounds easy.. but this wasnt a simple surface and image
problem. it is tough to explain without visuals, but you can check
out http://www.nefsc.noaa.gov/femad/ecosurvey/acoustics/ if you really
want to understand our data.)

rick - do you know Mike Jech?
david - i can understand your jealosy, but no need to be envious, we
can't all have big computers :) you might want to check out
comp.cold.shower

Report message to a moderator

Re: video cards [message #35256 is a reply to message #35142]

Thu, 22 May 2003 11:18

Rick Towler
Messages: 821
Registered: August 1998

Senior Member

"David Chevrier" wrote in message...

> I was wondering if anyone is using a 3D Labs Wildcat4 7210 video
> card. I wanted to know what people thought of this card and
> how they set their video options. I'm not setting the results I had
> hoped for.

You never do...

Just so we're on the same page, your Wildcat will only show it's teeth while
using hardware (openGL) rendering in object graphics. Matt's comment that
RSI designs the OG rendering system for maximum compatibility is right on.
Although I have no inside knowledge I would assume they use few if any GL
extensions so many of the features of mondern 3d display adapters go unused.
That being said, it doesn't mean you can't push a lot of pixels. The
Wildcat4 is an impressive card on paper. It should do quite well in IDL.

Performance depends a lot on what you are doing. Can you describe what you
are trying to do? How are you measuring performance?

> I know that some of this is due to both IDL memory
> limitations on 32-bit OS's (I'm using WinXP pro).

I don't think that this is a 32/64 bit issue as you are unlikely bumping
into the address limitations of the PC. In general, viz applications are
bottlenecked at the display adapter (depending on the vertex count/window
size).

> I write mostly object-oriented gui's for government acoustic
> data at NOAA.

What sort of acoustics?

Report message to a moderator

Re: video cards [message #35418 is a reply to message #35221]

Tue, 27 May 2003 14:04

David.Chevrier
Messages: 14
Registered: May 2003

Junior Member

>
> Just wait until you get that SM2000!

mult-beam! i have no idea how i'm gonna do visuals for that one! but
as we dont have any data yet, i dont need to worry about it... for now

>
> Are you working with Mike "Groundskeeper Willie" Jech? He didn't buy you
> that computer, did he? Looks like I need to upgrade...

"Groundskeeper Willie".... hmm, i'll have to ask him about that one.
he let me research and pick out all the parts. but i put this baby
together myself (and threw in a few hundred bucks of my own cash into
it too) its unbelieveable!! pretty good rig considering my degrees
in biotechnology and biology.

dave

Report message to a moderator

Previous Topic:	Re: How to Acces heap object.
Next Topic:	Minimization of deviations from multiple curve fits.

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Wed Oct 08 13:43:32 PDT 2025

Total time taken to generate the page: 0.00690 seconds