Very nice. Like always the vast majority of games will work just fine. Exceptions will likely exist, like Splinter Cell, but hard to quantify how many.

FX should be stronger with AA / AF eye candy than the GF4.

3DMark2000 default settings with P3 1000 (Coppermine): 6335 3D marks.

Compare to GF4 Ti4600:

3DMark2000 default settings with P3 1000 (Coppermine): 6713 3D marks.

Used 3dMark2001SE to test DX8 performance:

GF4 Ti4600: 6385
Quadro FX 500: 5123

So the Quadro is a more competitive DX7 card, only 5% off the Ti4600. But with DX8, it is about 20% off the Ti4600.

It does seem to be a very good card. I got some of these in a bulk lot about 1-2 years ago but haven't done much with them. I made sure they worked but I never took the time to compare them with other cards, so your test info is interesting. Looks like the FX500 would be close to a Ti4200 in DX8.
In the little testing I did it seemed that they have very good image quality and they also have top notch capacitors on them. Such are the perks of a Quadro card. The fans might be more reliable as well, but I don't know that for certain. 128-bit memory, no fear of ending up with a nerfed 64-bit card. I think the specs on this card are most closely equivalent to a 128-bit FX5500.
I'm surprised at how many of these are being sold pretty cheaply on eBay. If the performance meets one's need then I'd definitely recommend them.

This card was $11 NIB shipped. I bought two at that price and made an offer, so it's possible that I could have bought it even cheaper.

I don't have a Ti4200, but I benched a Ti4400 with both 3DMarks.

3DMark2000: 6534
3DMark2001: 6314

So I think you're right about the Quadro FX 500 probably being similar to the Ti4200 for DX7 games.

I was surprised by how close the Ti4400 was to the Ti4600 in both benchmarks. For DX8 games, the Quadro might be in the GF3 range. I have a Quadro DCC, so I'll test later.

Keep in mind that it is likely CPU limited somewhat. When the Ti 4xxx series was the biggest thing, the high end CPUs at the time would have been Thoroughbred A or B core Athlon XPs (which often hit upward of 2Ghz with overclocking), or a 2-2.53Ghz Pentium 4 Northwood.

The only incompatibility I know for sure is that it does not support VESA line frame buffer used by build engine games.

Ozzuneoj wrote:

Keep in mind that it is likely CPU limited somewhat. When the Ti 4xxx series was the biggest thing, the high end CPUs at the time would have been Thoroughbred A or B core Athlon XPs (which often hit upward of 2Ghz with overclocking), or a 2-2.53Ghz Pentium 4 Northwood.

Good point. I have a KT133A + Athlon XP 2400+ system running 98SE. I may bench on those too.

boxpressed wrote:

Ozzuneoj wrote:

Keep in mind that it is likely CPU limited somewhat. When the Ti 4xxx series was the biggest thing, the high end CPUs at the time would have been Thoroughbred A or B core Athlon XPs (which often hit upward of 2Ghz with overclocking), or a 2-2.53Ghz Pentium 4 Northwood.

Good point. I have a KT133A + Athlon XP 2400+ system running 98SE. I may bench on those too.

That should show a much larger performance gap between the cards, for sure. 😀

I had a Ti 4400 with a 1.33Ghz Athlon Thunderbird for a while, then later upgraded to an XP 1700+ and overclocked it beyond 2Ghz. Was a great combo until the poor Ti 4400 died and they replaced it with a garbage FX 5600 256MB (not even an Ultra). It was much slower than the 4400. I replaced that with a Tyan 9600 Pro and that was a phenomenal card... until the fan died... and they sent me a new fan, and the screws stripped out, making it impossible to replace the fan... so they sent me a new card! I still have that one. Those were turbulent times for me and PCs though... *shudder*

Come to think of it, I'm pretty sure I still have the retail box for my Visiontek Ti 4400 folded up and stashed away, along with ones from some old AMD and Intel CPUs from the 90s.

I tested the Quadro DCC, a DX8 card that is very similar in performance to a plain GF3. Here are the benchmarks in a more readable format:

1SE440BX-2 + P3 1000 CUMINE + 256mb RAM + Forceware 56.64
2
3                   3DMARK2000       3DMARK2001
4QUADRO DCC         6683             4875
5QUADRO FX 500      6335             5123
6GF4 TI4400         6534             6314
7GF4 TI4600         6713             6385

Has to be CPU limitation. They can't be that close.

Ozzuneoj wrote:

That should show a much larger performance gap between the cards, for sure. 😀

My tests bear out your point. The Ti4600 is now almost 40% faster than the Quadro FX 500 for DX7 and DX8.

What's interesting is how well the Quadro DCC did: over 20% better for DX7 and almost as good for DX8.

All in all, I'd say that the Quadro FX 500 is an inexpensive alternative to cards like the GF3 Ti500 / GF4 Ti4200 if you're running a P3. Once you go to a P4 or Athlon XP, you'll want a beefier card.

(My system crashed running 3DMark2001 using the 56.64 and 61.76 drivers, so I used 45.23.)

1K7VTA Pro + Athlon XP 2400+ + 256mb RAM + 45.23
2
3                   3DMARK2000       3DMARK2001
4QUADRO DCC         9032             6369
5QUADRO FX 500      7352             6540
6GF4 TI4400         10309            8645
7GF4 TI4600         10396            8933

It would be interesting to see how the rare Quadro FX 600 fits in to these tests.