|
Elsa
Gladiac 921DVI GeForce3 Ti500
|
|
Author : Wayne
Date : 30th December 2001
|
3DVelocity would like to
thank Elsa
for their help and courtesy in providing this graphics card
for review.
Benchmarks
For reference, I will be
comparing the Gladiac to the Creative Labs 3d Blaster GeForce3
Ti500 I reviewed just before Christmas. Both cards closely follow
the reference design which usually means they benefit from NVIDIA's
extensive reliability testing. 21.85 drivers were used throughout.
Where better to start than
with MadOnion.com's 3DMark 2001. I'm not sure if I'm becoming
paranoid (or cynical) but I ran the tests on my own copy of
3DMark rather than use the supplied copy. I guess the ATi Quack/Quake
debacle has gnarled my trust.
3DMark 2001 :

I don't think anything I say will make the results
any clearer than they are when you look at the graph. Elsa takes
the edge at all bar 1600x1200, if you can call it an edge. The
differences are well within the allowances for error so the
performance is essentially the same.
VulpineGL :
VulpineGL is a prime example of demanding OpenGL
with high polygon models and large distance maps. It places
heavy demands on your GPU and really sorts the men from the
boys. I have used the screenshot from the Creative review to
show the settings used and I ran the test four times, twice
with sound enabled and twice with it disabled before averaging
the results. Just so you know, enabling sound made a difference
of between 0.1 FPS and nothing at all on my Audigy.


Before you get too excited, the massive difference
that appears in the above graph actually amounts to no more
than 0.2 FPS. Again, allowing for deviations in test conditions
there's nothing between them.
GL Excess :
If you've seen my other reviews you'll know how
fond I am of this OpenGL benchmark. The download is surprisingly
compact considering the visual quality and the results seem
very consistent.
|
Legend
|
|
FILL
RATE Tests
|
|
POLYGON
COUNT Tests
|
|
VRAM
Tests
|
|
CPU/FPU
Tests
|
| Creative 3D Blaster |
5163
|
7176
|
3398
|
5573
|
| Elsa Gladiac |
5170
|
7181
|
3387
|
5577
|
Again we see almost indistinguishable scores, though even
allowing for errors the Elsa does seem to be scraping enough
victories now to suggest an ever so slightly more efficient
design. Maybe a few more results will help form a better picture.
Page 9 - More Performance
