You can use the V5 PCI in the newest mainboards but not the AGP version because it only supports 3,3V.
If you have many PCI cards in your system the AGP will be faster.

@grunw,

I have both a 5500 agp & pci, the performance difference is marginal at best, through testing on the same system (xp2400+) I have found the agp to be just slighty faster "just". You can get slighty more performance also with a agp version if you'r overclocking your processor and the agp bus speed gets increased, mine is at 83 mhz versus the standard 66 mhz.

The only real difference between the agp and pci versions are the differences in the bus speeds (pci = 33 mhz versus agp = 66 mhz)

Hope this helps

Voodoo5 PCI is PCI66 device, so it runs at 66MHz on PCI 2.2 motherboard.

Same way as PCI-eXpress is now a marketing strategy for ATi/nVidia/whoever.

I mean, the speed of the bus itself only affects data transaction,
now, when was the last time that you've seen a performance increase when going from AGP1x to even AGP8x?
Or when using fastwrites?

Truth is never, gfx cards nowdays almost never do swapping anyway, they don't need those high speed buses.
Heck, the Hercules 3D Prophet 8500 64mb that I have, doesn't even swap in DeusEx2, and in high details - that game would even make your 9800pro beg for mercy.
So anything above AGP1x is pure marketing hype.

I mean, 10gbit ethernet might need PCI-eXpress, uh.. 50 megapixel video cameras might need PCI-eXpress, maybe even some extremely-high-end sound cards.
But video cards? pft

Well, since all Voodoo4/5's only effectively communicate over 33MHz 32bit PCI Bus (never seen any of my Motherboards actually activate the PCI66 option), it is half the speed of the 66MHz 1xAGP version and a quarter of the 2xAGP version.

Since that's only valid for actual Transfers via the AGP Bus, and the Command FIFO not using it excessively, the ~10% (can be almost 0% in other applications) advantage is normal.

Newer Games (e.g. UT2003) should make the biggest difference, as they swamp the Bus with T&L Data for the Cards to render.

Well, that's Tom as we know it

Anyway, it largely depends on the Application you throw at it.

I mean, run Unreal Tournament on 2x/4xAGP vs. 8xAGP Cards, and you'll see those 1-2% (at best).

Do the same using the (extreme by all means) CodeCreatures benchmark, and you can see upto 60% in isolated Cases.

The trick is to force the Card into AGP Texturing, that's it.

But still, the usefulness of 8x AGP is (today) mainly limited to getting T&L data and other misc. stuff over the Bus to the Card quick.
AGP Texturing is really not used alot (if at all), since it drops performance way too low.

But if you want to see 8x AGP at it's best, use a 64MB Card, set it to 2xAGP and run UT2003 in maximum Texture Detail (requires manual Engine flush to enable true 256MB Textures); 128MB (default maximum) should so as well however.
Then, repeat procedure on 8xAGP.

(Problem here is finding a 8xAGP capable Card that actually still supports 2xAGP)

----------------
The difference from 4xAGP to 8xAGP in turn, is neglectable, simply because 4xAGP already almost exceeds most system's total available "spare" RAM bandwidth (which is naturally shared among other Tasks, e.g. PCI Bus & CPU).
So 8xAGP would just bottleneck the RAM even more, yielding either no reasonable performance increase (worst case), or still less than expected...

It's just hard to find the right combination of Cards and Applications that posess enough Textures to force AGP Texturing, and provide a Platform that can spare sufficient RAM bandwidth for the 8xAGP to actually use as advantage over 4xAGP.
-----------
But by all practical means, I agree :
We don't really need 8x AGP as of now.
And when the time comes we could actually need it, we'll find the performance drop being probably too significant to be useful.

Even that System could bottleneck it enough to reduce 8xAGP efficiency, trust me

The basic Problem is :
Applications that ship sufficient Textures to overload modern 128MB Cards (and potentially show the use or limitations of AGP Texturing) usually are very CPU intensive as well, and cause a very high Memory bandwidth demand all by itself already.

A P4EE with 2Gigs of RAM is already limiting itself, since it's forced into using slow PAE, which causes all P4's to take a notable hit (can lose upto 30% performance by that alone)

With a completely synthetic, pure AGP Texturing benchmark, one could even achieve much higher difference in scores, but so far I've never seen one.
(not very useful in general anyway, that's probably the reason none exists)