cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Did you get called away in the middle of writing a post? Don't worry you can find your unfinished post later in the Drafts section of your profile page. X

Video Card performance test OLEF or PassMark

amedina
1-Newbie

Video Card performance test OLEF or PassMark

I was using the OLEF test to perform some tests here when I stumbled upon
this other test site:



This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.
7 REPLIES 7

"

> The test is really when you have 5 windows open at the same time - if you
> can live with one, two at most - then the GTX will be okay
>
> "
>

Ok, But can you confidently say that a GTX580 will not open as many windows
as an FX3500 from several years ago? the FX3500 has 256MB GDDR3 while the
GTX580 has 1536 MB GDDR5. not only is it 6 times more ram, but its also
almost twice as fast.

SO the question about this statement is, what would the card be doing with
so much ram such that the FX3500 could open more windows.

I've seen before with an old ATi card 64mb that pro/e would send a dialog
saying "please close some windows" when the card could not handle more. So I
am aware that if the ram is low, you cannot get as many windows.

So is a top of the line consumer card much much better than a professional
quadro card? I really want to hear from someone who went from a quadro to a
GTX and said, "WOW that's fast"

I want to keep hearing about the video card devate. New consumer level VS
older Professional level, which is best?

thanks for the reply

I'm looking at the Gigabytes per second only as a way to gage speed. (is
bandwidth related to pipelines per pixel? I'm thinking pipelines per pixel
is a subset of the total bandwidth bottle neck, so bandwidth should that
into account?)

For example, the:
Q FX 380 can do 12GB/s
Q FX 580 can do 25GB/s
Q FX 3500 can do 42GB/s
Geforce GTX can do 192GB/s


how may windows can I open? this is a memory question
Q FX 380 has 256mb
Q FX 580 has 512mb
Q FX 3500 has 256mb
Geforce GTX has 1536mb




"The test is really when you have 5 windows open at the same time - if you
can live with one, two at most - then the GTX will be okay"

"I have run pro on many of the GT, GTX gaming cards with great success. My
suggestion is to buy the card from a local store that allows returns, and
if it does not work return it within 30 days."

"

The problem is what the card is ‘tuned’ for. Consumer cards are designed to
run Direct-X. Professional cards are designed to run Open-GL.

CAD systems use Open-GL.

When you use a consumer card with a CAD system, the CPU has to convert all
the graphics commands to Direct-X graphics calls BEFORE sending them to the
graphics card. This slows your system performance.

When using a professional card, the CAD application sends the Open-GL
graphic calls directly to the graphics card.

"

I cannot say much about the new consumer grade cards, but I recently
migrated from a GeForce 7800 GT to a Quadro FX 4500. I was pretty happy
with the GeForce, but I stumbled on a great eBay deal for the Quadro and
"upgraded". I ran PassMark to gauge the performance bump and the raw
numbers were disappointing (most of the 2D scores, including the composite
score, were actually very slightly LOWER on the Quadro with 512 Mb ram vs
the GeForce with 256 Mb. on the 3D tests, the Quadro scored higher in
everything, but only by ~5% or so.)



Now that I am revisiting the numbers, I am really disappointed.



However. I pulled up a big, busy assembly (WF2) that I am familiar with and
I have to say that my disappointment dissolved. I am very happy with my
real-life experience with the Quadro - it prehighlighted and selected very
busy components in a blink of an eye, where the 7800 GT would take a bit
longer so you could watch the selection progress across the part. With
components selected, the spin rate was noticeably smoother using the Quadro.
I cannot explain why the PassMark numbers don't tell the story, but I am
happy with my purchase. I would love to throw in a GTX to gauge the
difference again. Maybe someday I will spring for one.



My only thoughts on the numbers not mirroring my experience is that possibly
the PassMark performance test is designed for gaming and not CAD? As well
as the concept that the GeForce cards are designed for gaming and Quadro's
are designed for CAD. Just a guess. If you have another opinion, I would
love to hear it.



Alfonso, Thanks for the link and the thread.



PassMark is ranking the GeForce 7800 GT at 197

the Quadro FX 4500 is ranked at 201



I guess that explains / supports my findings - Wouldn't it be nice if they
had a third column in the rankings: price$



-Nate


dgschaefer
21-Topaz II
(To:amedina)

Pro/E uses the graphics card differently than games.  For game, card
memory is important, but I understand that with Pro/E it's almost
irrelevant.  For general model shading, Pro/E doesn't use that memory at
all.  It may for rendering or mapping decals, I'm not sure (and I'm sure
someone can correct me if I'm remembering this wrong)



The GeForce gamer cards fall down when multiple windows are opened.
Why? I dunno, but we has a shared laptop here a couple of years ago with
a GeForce card and open more than 3 or 4 windows and Pro/E would slow to
a crawl and sometimes the entire window would turn gray.  Close that
last window, and all was right again.



My guess is that test is measuring the things important to gaming, not
CAD.

Doug Schaefer
--
Doug Schaefer | Experienced Mechanical Design Engineer
LinkedIn
StephenW
23-Emerald II
(To:amedina)

I don't have any experience with gaming cards but about 1-1/2 yrs ago we upgraded the video card in my computer from a Quadro FX1800 to a Quadro FX4500. I OCUS benchmarked before and after so I could get a quantitative value for the improvement.

I only did the benchmark once for each card so if something happened to run in the background during one of the test, my numbers wouldn't flush that anomaly out, but I did disable virus protection for the test.

The first column is the FX1800, second column is the FX4500, third is percent change. Numbers are in seconds so LOWER IS BETTER.
The final results are shown here. You can interpret them as you see please.


[cid:image002.jpg@01CBDE33.108178D0]

I have used GeForce 7800 video gaming cards for several years with absolutely no problems whatsoever. I can have 10 Pro/E windows open without a hitch.

Bob

Yesterday we took our several OCUS5 tests and compared them to the PassMark
test and they correlate very closely. We read the test info and apparently
their test only tests for graphics. So that is what we compared it to, just
graphics. It turns out that an IBM E20 with a Quadro FX1800 would perform on
the OCUS5 test, 1.7 times better by using a GTX580 card. My coworker looked
at a comparison between an IBMM90 using an FX3500 and the same computer
using an FX380LP. The numbers for that comparison were the same between the
OCUS test and the PassMark test. (graphics only).

I'm leaving this topic open, If anyone is brave enough to go to frys and get
a GTX580 to try it out, let me know. Just keep that receipt just in case.
The OCUS5 test does have a few GeForce cards in there, but so far have not
seen the newer GTX represented. I saw a video that made me wonder a few
days ago:

Top Tags