Better to look at GeForce4 MX than a 7500. Faster and better game compatibility.

7500 is certainly an interesting improvement to the original Radeon though, and a bit of a curiosity. It has various bug fixes, including more anisotropic settings, and asynchronous GPU:RAM clock support. GF2U is a bit out of its reach though. It's more a GF2 GTS beater.

noshutdown wrote:

thats only about theoretical specs. in real life, all nvidia cards prior to geforce3 suffer from unoptimized z-buffer usage, overdrawing of hidden surfaces, and less efficient RAM controller, which wasted most of its fillrate and memory bandwidth and lead to a lower efficiency.

Nvidia later implemented significant hidden surface removal at driver level. Care to elaborate why was the memory controller worse then that of Radeon?

noshutdown wrote:

for ati, these problem were addressed as early as in the rage128 cards...

I guess you mean only true color performance here.

noshutdown wrote:

thats only about theoretical specs. in real life, all nvidia cards prior to geforce3 suffer from unoptimized z-buffer usage, ove […]
Show full quote

sliderider wrote:

A GeForce2 Ultra should be a lot faster than a Radeon 7500. It should be closer to a Radeon 8500LE. […]
Show full quote

A GeForce2 Ultra should be a lot faster than a Radeon 7500. It should be closer to a Radeon 8500LE.

GF2U vs Radeon 7500

http://www.gpureview.com/show_cards.php?card1=124&card2=239

GF2U vs Radeon 8500LE

http://www.gpureview.com/show_cards.php?card1=124&card2=23

Amd don't forget that the Radeon drivers that cover cards from this time period suck and the nVidia ones rock so the GF2U would be even faster taking drivers into account.

thats only about theoretical specs. in real life, all nvidia cards prior to geforce3 suffer from unoptimized z-buffer usage, overdrawing of hidden surfaces, and less efficient RAM controller, which wasted most of its fillrate and memory bandwidth and lead to a lower efficiency.

for ati, these problem were addressed as early as in the rage128 cards, but for nvidia they persisted until geforce3, and the solution was advertised as "lightspeed memory architecture" with geforce3.

and yes, while ati's drivers do suck, the 7500 is still able to edge out the GF2U in real life performance. its the same case as the mx440, whose theoretical specs were far below the GF2U(but with its defects addressed), can edge out the 7500.

I have to disagree and so does Anand

In only 2 of the games tested does the Radeon 7500 run faster than the GeForce2 Pro but the couple of frames that you pick up in those games probably won't be noticeable in actual gameplay.

Also, the test is using a GeForce2 Pro and not the faster GeForce2 Ultra so the difference in the clock rate should pull the GF2U ahead of the Radeon 7500 in the games where it loses even with a less efficient architecture and even further ahead of the 7500 in the games where it wins. The 7500 is faster in theoretical tests like 3DMark 2001, but in actual games the GeForce2 would be a better card.

Something else I should point out, notice how the GeForce2 Pro sticks close to the GeForce 3 in the Unreal Tournament and Counterstrike tests. That says to me that those games are probably CPU limited by the test configuration and with a faster CPU that the GF2 has more potential than the 7500.

swaaye wrote:

Better to look at GeForce4 MX than a 7500. Faster and better game compatibility.

But what's it like when it comes to DOS games? Also, my motherboard is AGP spec 1.0 (1x, 2x) and runs at 3.3V... I don't think it'll play nice with an AGP 3.0 8x card 😢 .

It's great. Every Nvidia card is great for DOS, SVGA games work like a charm too. I've played Mechwarrior 2 at 1024x768 on a GF4 Ti 4200 and even a GF6600GT 😜

sliderider wrote:

I have to disagree and so does Anand In only 2 of the games tested does the Radeon 7500 run faster than the GeForce2 Pro but th […]
Show full quote

I have to disagree and so does Anand
In only 2 of the games tested does the Radeon 7500 run faster than the GeForce2 Pro but the couple of frames that you pick up in those games probably won't be noticeable in actual gameplay.

Also, the test is using a GeForce2 Pro and not the faster GeForce2 Ultra so the difference in the clock rate should pull the GF2U ahead of the Radeon 7500 in the games where it loses even with a less efficient architecture and even further ahead of the 7500 in the games where it wins. The 7500 is faster in theoretical tests like 3DMark 2001, but in actual games the GeForce2 would be a better card.

Something else I should point out, notice how the GeForce2 Pro sticks close to the GeForce 3 in the Unreal Tournament and Counterstrike tests. That says to me that those games are probably CPU limited by the test configuration and with a faster CPU that the GF2 has more potential than the 7500.

this appears to be an early test, and all i can see is how poor ati's early drivers were, with 8500 even losing to GF3standard by distance. this is just fucking stupid because with later mature drivers, 8500 can actually edge out the GF3ti500(this is still not a good score though, as 8500 actually had the same theoretical specs as the GF4ti4400).
why not look at this test in 2002? with 7500 beating GF2U in 4 tests out of 5:
http://www.tomshardware.com/reviews/vga-charts-ii,566-3.html

PowerPie5000 wrote:

swaaye wrote:

Better to look at GeForce4 MX than a 7500. Faster and better game compatibility.

But what's it like when it comes to DOS games? Also, my motherboard is AGP spec 1.0 (1x, 2x) and runs at 3.3V... I don't think it'll play nice with an AGP 3.0 8x card 😢 .

MX440 is an agp4x card and can run at 2x, there won't be compatibility problems, and dos performance is also fine.

I'm now looking at 2 cards... The unknown 64MB GF2 in the photo i posted earlier (could be a Visiontek/OEM GF2 Ultra?) and there's also a confirmed Visiontek 64MB GF2 GTS. Visiontek wasn't very well known back then, but i have no issues regarding their quality as they actually built the GF2 cards for Elsa (which we all know are good quality 😀).

Anyway, would i notice any difference between a GF2 GTS and Ultra with my setup? I'm just hoping the image quality is acceptable (not expecting it to be as clean and colourful as my G400 Max... But it would be nice 😀). Both cards are keyed for 3.3V and 1.5V AGP slots, so should work with my Intel SE440BX2 motherboard (at 2x speed). Just need to check my PSU will be enough... It's a Seasonic unit which is either 200W or 250W (can't quite remember). I also have a spare 300W PSU with ok specs, but it's a cheap brand (Mercury - which also used to manufacture nVidia cards).

noshutdown wrote:

this appears to be an early test, and all i can see is how poor ati's early drivers were, with 8500 even losing to GF3standard b […]
Show full quote

sliderider wrote:

I have to disagree and so does Anand In only 2 of the games tested does the Radeon 7500 run faster than the GeForce2 Pro but th […]
Show full quote

I have to disagree and so does Anand
In only 2 of the games tested does the Radeon 7500 run faster than the GeForce2 Pro but the couple of frames that you pick up in those games probably won't be noticeable in actual gameplay.

Also, the test is using a GeForce2 Pro and not the faster GeForce2 Ultra so the difference in the clock rate should pull the GF2U ahead of the Radeon 7500 in the games where it loses even with a less efficient architecture and even further ahead of the 7500 in the games where it wins. The 7500 is faster in theoretical tests like 3DMark 2001, but in actual games the GeForce2 would be a better card.

Something else I should point out, notice how the GeForce2 Pro sticks close to the GeForce 3 in the Unreal Tournament and Counterstrike tests. That says to me that those games are probably CPU limited by the test configuration and with a faster CPU that the GF2 has more potential than the 7500.

this appears to be an early test, and all i can see is how poor ati's early drivers were, with 8500 even losing to GF3standard by distance. this is just fucking stupid because with later mature drivers, 8500 can actually edge out the GF3ti500(this is still not a good score though, as 8500 actually had the same theoretical specs as the GF4ti4400).
why not look at this test in 2002? with 7500 beating GF2U in 4 tests out of 5:
http://www.tomshardware.com/reviews/vga-charts-ii,566-3.html

Here's another test from Xbitlabs

http://www.xbitlabs.com/articles/graphics/dis … adeon-7500.html

This one puts both the GF2 Pro and GF2 Ti against the Radeon 7500. The 7500 excels in the theoretical 3DMark 2001SE but in actual games the GF2 still comes out looking like the better card overall and the GF2 Ultra can only add to the scores achieved by the Ti.

@noshutdown; those Dungeon Siege benchmarks make me laugh. How can a Radeon 9000 Pro possibly outperform a Radeon 8500? What's even funnier is that the 7500 has a higher framerate on the AMD system than the 8500 cards 😁

I might just avoid Radeon cards... It seems a bit of hassle just to change the heatsink & fan if they die (which is inevitable), Unless i buy a Hercules Radeon, but i also heard early 3rd party Radeons can be a bit iffy 😒. It seems most (if not all) GF2 cards have push-pin heatsinks and standard fans that can easily be removed and replaced 😀.... I've knackered graphics cards before when trying to pry off a heatsink that's stuck on with epoxy 😠 .

What kind of PSU will a GF2 Ultra or GTS need? Is the Radeon 7500 more power hungry?

PowerPie5000 wrote:

I might just avoid Radeon cards... It seems a bit of hassle just to change the heatsink & fan if they die (which is inevitable), Unless i buy a Hercules Radeon, but i also heard early 3rd party Radeons can be a bit iffy 😒. It seems most (if not all) GF2 cards have push-pin heatsinks and standard fans that can easily be removed and replaced 😀.... I've knackered graphics cards before when trying to pry off a heatsink that's stuck on with epoxy 😠 .

What kind of PSU will a GF2 Ultra or GTS need? Is the Radeon 7500 more power hungry?

I can't find many reviews from back then that actually worried about power consumption. I don't think that became a big issue until much later. The GF2U and the Radeon 7500 don't have external power connectors so the draw can't be any higher than what an AGP slot can handle.

sliderider wrote:

PowerPie5000 wrote:

I might just avoid Radeon cards... It seems a bit of hassle just to change the heatsink & fan if they die (which is inevitable), Unless i buy a Hercules Radeon, but i also heard early 3rd party Radeons can be a bit iffy 😒. It seems most (if not all) GF2 cards have push-pin heatsinks and standard fans that can easily be removed and replaced 😀.... I've knackered graphics cards before when trying to pry off a heatsink that's stuck on with epoxy 😠 .

What kind of PSU will a GF2 Ultra or GTS need? Is the Radeon 7500 more power hungry?

I can't find many reviews from back then that actually worried about power consumption. I don't think that became a big issue until much later. The GF2U and the Radeon 7500 don't have external power connectors so the draw can't be any higher than what an AGP slot can handle.

I'm just hoping my old AGP 1.0 (2x) slot can output enough power for these cards... My PSU is also only 200W or 250W (Seasonic).

NitroX infinity wrote:

@noshutdown; those Dungeon Siege benchmarks make me laugh. How can a Radeon 9000 Pro possibly outperform a Radeon 8500? What's even funnier is that the 7500 has a higher framerate on the AMD system than the 8500 cards 😁

about that test, i would say that its cpu-bound, so ti4600, r9700pro and geforce3 all have similar score.
in such a case, the driver would play an important role on how much cpu power it consumes, and thats why 8500 and 9000(cut down of 8500) have almost identical score, while they are both a lot slower than the ti200 which is a lower-end card.

sliderider wrote:

Here's another test from Xbitlabs

http://www.xbitlabs.com/articles/graphics/dis … adeon-7500.html

This one puts both the GF2 Pro and GF2 Ti against the Radeon 7500. The 7500 excels in the theoretical 3DMark 2001SE but in actual games the GF2 still comes out looking like the better card overall and the GF2 Ultra can only add to the scores achieved by the Ti.

as you can see, this test was also published on early 2002, which is not far from 8500/7500's release. and yes, ati's drivers take almost an eternity to mature and to make most performance out of it.

noshutdown wrote:

...and thats why 8500 and 9000(cut down of 8500) have almost identical score, while they are both a lot slower than the ti200 which is a lower-end card.

FYI Radeon 8500(LE) was almost at half a price of GF3ti when it came out. I know because I wanted a cheap alternative to the 'legend'. ti200 was never a lower-end card. Also, the 8500 was never a great card - only 'decent'. I had one and I was never really satisfied; it was always either too slow, or too lousy. Imagine my jaw-dropping when I finally moved on to radeon 9500 (and unlocked to 9700); it was as much of a performance jump as it was from K6/2 500 to Duron 750 (OC @ 1Ghz).

...but to say that GF4mx or even GF2 could match up to radeon 8500 is also overstatement. I had a crazy overclockable Abit Siluro MX 440 that scored almost 9k in 3dmark01 (the radeon 8500 scored about 10.5k in the same machine), but it could never match the performance of radeon 8500. Really. Especially in DX 8 games. 😎

NitroX infinity wrote:

@noshutdown; those Dungeon Siege benchmarks make me laugh. How can a Radeon 9000 Pro possibly outperform a Radeon 8500? What's even funnier is that the 7500 has a higher framerate on the AMD system than the 8500 cards 😁

Probably driver problem with particular game, but for example RV200 had faster pixel shader then R200.

For it's price I loved my 8500LE.

I was a 8500 guy too. It sounds awesome on paper, and I got one for only $90, but that was one bugged card. It has its share of hardware problems, like even though it should be faster than GF4 in Doom3 because of PS1.4, it is slower because of some internal quirks. Truform is mostly ridiculous. It entirely lacks MSAA and its SSAA is surprisingly disappointing even though it's said to be rotated grid of some sort. Its anisotropic filtering is fast but that's because it's very limited in quality. Backward D3D compatibility is poor. It is ok for most D3D7 and 8 games, and Quake engines work fine.

You guys really don't want ATI for your old games. But maybe you just want to mess with everything anyway.

RV250 is apparently more efficient because otherwise it should not be able to keep up with R200. It lacks the Truform hardware, somehow emulating it. It has only one vertex shader whereas R200 supposedly has 2 vertex processors.

I've decided not to bother with a Radeon... Just trying to find out if a Geforce 2 GTS/Ultra or even a Geforce 3 will be fine with my 3.3V AGP 1.0 slot (supports only 1x and 2x speeds). Does a Geforce 3 have better image quality than a Geforce 2? I'll probably change the CPU to a PIII 850Mhz (fastest the mobo will take) if i get a GF3 Ti200.

I wish i could find info on power consumption for all these cards... I'm guessing a GF2 Ultra would draw more power from the AGP than a Geforce 3 Ti200? Or is it the other way?

Yeah I said you can run anything from GeForce FX and older on AGP 3.3v like a 440BX. I've played Doom3 on a 440BX with a FX 5900 Ultra.

GeForce 3 is definitely better than GeForce 2 for image quality. GeForce FX is better than GeForce 3/4.

Power consumption isn't really something to worry about. If you want really low power consumption for some reason, look for cards with little fanless heatsinks or bare chips (some MX cards).