Voodoo 5 5500 review

Hilbert Hagedoorn · 08-05-2000, 11:36 | posts: 16,989 | Location: Guru3D testlab

Please post your comments here.

08-09-2000, 23:49 | posts: n/a

I am really sad that you guys can not get an objective review. I mean in your first paragraph you state:"Theoretically you can place up-to 32 VSA-100 chips on a videocard. Considering that each chip uses 15 Watts and needs 32Mb memory we might as well buy an external generator next to the PC for such a card LOL!"You happen to forget to mention that that could buy you up to 3.2 GPs and 40 GBs memory bandwidth. Pretty big numbers. You also say that 4xFSAA is too much of a performance hit. And your right for fill rate limited games. But there are many games that are not fill rate limited...Why was this not mentioned???Also in your Q3 bench marks even though the V5 loses every test, the delta is less than 5fps. 5 fps is awfully close. Blurry images? No mention of the LOD setting :-( Another comment that dose not make much sense:"What also is interesting to see is the performance of the GeForce 256 (SDR Memory I must add)"The V5 uses SDR memory. When you compare the GF2 200Mhz DDR, to the V5 with its SDR 166Mhz memory you still have to be impressed that the difference is only 5fps or less. Can we say memory bandwidth issue on GF2? The GF2 has a higher pixel fill rate and yet dose not pull into a larger lead when the card becomes heavily loaded...Then you knock them for the power connector. Well in case you have not heard there are many reports of issues with GF2 on the APG. Now the fault may lie with the mother board and it poor power via the APG. But the V5 has a power connector to avoid this, which so far seems to have much more stable results, and they get dinged for it?? Dose not make much sense....Price?? Go to price watch and you will see that the V5 goes for $251. The lowest 32Mb GF2 is $243, the lowest GF2 with 64mb memory $299.No UT benchmarks? I know UT is not a good benchmark. But it is the 3rd most popular game played on-line with many titles based off its engine (DNF, Dexus, Kligon Academy, etc) Wont people want to see how the card runs that game??? Looking at the other reviews you gave all of the GF2 cards higher scores when they have the same issues How did you come up with the general stability & Ease of Use?You also made several good points. I am not saying this reveiw was wrong. Just one sided. Many key points were left out and some were a bit off. Jb

Hilbert Hagedoorn · 08-10-2000, 00:47 | posts: 16,989 | Location: Guru3D testlab

Ah, a true 3dfx fan I guess:>You happen to forget to mention that that >could buy you up to 3.2 GPs and 40 GBs >memory bandwidth. Pretty big numbers. Pretty big numbers indeed, realistic ? Not really, it would use 15x32= 480 Watts !Also, each VSA-100 uses 4*8MB SDRAM modules times 32 = 128 SDRAM (1024MB) chips on one PCB ? Not possible. Again, the theory is nice, the numbers are good but it's not likely .. Why do you think the V5 6000 still has not been released ? It surely isn't marketing.>You also say that 4xFSAA is too much of a >performance hit. And your right for fill >rate limited games. But there are many >games that are not fill rate limited...Why >was this not mentioned???We test cards for current and upcoming releases. FSAA was never ment to be designed for fillrate limited games. That's a rediculous assumption. Anyway, I put your statement to the test by using NFS Porche Unleashed (which is very fillrate limited). It still is not playable in 4x FSAA and I really do not play Pac Man anymore (J/K!).>Also in your Q3 bench marks even though the >V5 loses every test, the delta is less than >5fps. 5 fps is awfully close. True ! However, I just tested it on a 700 Mhz Thunderbird, FPS differences vary up-to 20 FPS in Quake III (in favor of GeForce2). The new benchmarks will be included in an update soon. It seems that Intel's MTH bugged i820 mainboard limited the GeForce2 in high memory bandwidth. >Blurry images? No mention of the LOD setting :-(Blurry Images that are produced by FSAA <IMG SRC="http://guru3d.com/ubb/smile.gif">>Now the fault may lie with the mother board >and it poor power via the APG. But the V5 >has a power connector to avoid this, which >so far seems to have much Although some mainboards reported power supply issues in the past I believe that the primary power issue needs to be solved with a good power supply. If you have an Athlon based system with a voodoo5 5500 in a mini-tower (200 Watts PS) then you might be in for some trouble, wether you draw the power from the PS or either the mainboard the result remains the same, although pulling 30 watts from your mainboard would be ludicrous and would make the mainboard's AGP highly unstable.My point, (BTW I'm not a pro NVIDIA freak or something, these are my personal findings/opinions) Look what the GF2/MX can do with less than 10 Watts and what the Voodoo5 5500 does with 30 Watts. I'm strongly against Multi-processing videocards and that consume a lot of power, I really don't care if it's a ATI, Matrox, 3dfx or NVidia videocard.>Price?? Go to price watch and you will see >that the V5 goes for $251. The lowest 32Mb >GF2 is $243, the lowest GF2 with 64mb >memory $299.Jep, however, in terms of speed the Voodoo5 5500 is on par with GeForce2 MX, that's where you'll find the most bang for your bucks.>No UT benchmarks? I know UT is not a good >benchmark. But it is the 3rd most popular >game played on-line with many titles based >off its engine (DNF, Dexus, Kligon Academy, >etc) Wont people want to see how the card >runs that game???UT is not a reliable benchmark.>How did you come up with the general >stability & Ease of Use?Gen. Stab - The Voodoo5 5500 crashed on some occasions.Ease of use - The voltage connector is simply a drag <IMG SRC="http://guru3d.com/ubb/smile.gif">>Then you knock them for the power connector. Knock them ? These are my words "Installation of the videocard was somewhat strange."The fact however remains that here in the Netherlands we have to pay 400 bucks for a videocard which is simply not worth that kind of money.JB, thanks for your comments, I really do appreciate constructive feedback. I stand behind my review and will back it up in any way I can. However, I can not agree with a lot of what you are saying, but that's ok!, those are your findings/opinions and what is noted in the review are mine. One thing that disturbes me though is that you call this review 'not objective' ... while the article in all ways objective is <IMG SRC="http://guru3d.com/ubb/smile.gif">

08-10-2000, 04:20 | posts: n/a

Not another 3dmark 2000, Quake 3:Arena review. I wish there were better benchmarks than using one game that I couldn't care less of and a synthetic benchmark. I just bought a Voodoo 5 5500, and it whips through all the games I like to play (UT, Tribes, Ultima 9, Rogue Spear (4X FSAA) and Diablo 2). If any of your readers like the games above, go for the Voodoo 5 5500 as they all have Glide renderers except Rogue Spear. BTW, the Geforce 2 GTS and Geforce 1's T&L will be obsoleted by DirectX 8.0, they only support 2 matrix blending instead of 4. If you really want to be future proof today, purchase an ATI Radeon as it supports 4 matrix blending in it's T&L.

08-10-2000, 08:19 | posts: n/a

"Ah, a true 3dfx fan I guess:"Hmm maybe. Maybe not. I am a person that like to fight for the underdogs. Plus I have seen so many people pick a part the V5 that maybe I have gotten full of enough BS on both side. Of course I bought a V5 a month ago. I was really close to a GF2. But I held back for Glide and better FSAA. "Pretty big numbers indeed, realistic ? Not really, it would use 15x32= 480 Watts !"True but that was for a high end, ok really high end graphic board for only a small marked. Since most Workstation High end GL cards that use mutilple processors and if you look at the latest edition of PC Magazine you will see that they often need to run on a APGPRO 110 bus. I did not realize but I get there are some APG buses that can souce 110 watts of power. Not much to discuss here as this reveiw was not about the Ultra high end workstation multi-processor cards. " Why do you think the V5 6000 still has not been released ? It surely isn't marketing."I have no clue. Maybe the 3DFX PR people are right and we will see it soon...Maybe not.... " Anyway, I put your statement to the test by using NFS Porche Unleashed (which is very fillrate limited). It still is not playable in 4x FSAA and I really do not play Pac Man anymore (J/K!)."As others have said there are plenetly of games that ARE NOT FILL RATE LIMITED. CS is a game I play at 800x600 4X. Looks amazing and silkly smooth. NO visible frame rate loss. I dont play at higher res as the targets and cross hairs become much smaller = harder to hit...I mean the accuracy of the latest realses makes it hard enough to get a hit. IMHO if you like fill rate limited games, then GF2 is your card. Free Space looks awesome at 1024x768 x4 FSAA."Blurry Images that are produced by FSAA"I agree, but with a small tweak to the LOD you get that sharp image back. the first time I saw Q3 at 800x600 4X FSAA with LOD = -2 (or -8 depending on your driver version) it was awesome. Plus at Higher Res you still have pixel poping and texture crawling. Look at the Unreal flybys and you will see pixel popping. That goes away with FSAA but dose not with high res (although it is harder to see)."I'm strongly against Multi-processing videocards and that consume a lot of power, I really don't care if it's a ATI, Matrox, 3dfx or NVidia videocard."??? I agree. Also you have to respect the benfits of SMP systems. If designed correcly they can give you more preformance than a signal processor. As with all engineering there is always a trade off. I did think 3DFX did a good thing with the VSA. The wanted to have multiple cards to fit the low, med and High. Since they have to worry about inventory cost and nVida dose not, then there is some merit of teiring the VSA products. Step back and look at the VSA degin and you see that it has some merits. Besides costs/inverntory look at Memory Bandwithd. Here the V5 has the same bandwidth as teh GF2 using a lower clocked and slower SDR than the GF2 DDR. Again there are alwaystrade offs... "UT is not a reliable benchmark." There is a really good discussion over at beyond3d about benchmarking. On one half you have vaild points about how Q3 and 3D Mark test the limits of a graphics cards. But on the other hand if a card A preforms better than card B in the same game, then I feel it is a vaild bench mark. Of course your not only benchmarking the Vid but Drivers, Mem, CPU, ect... And since you need all of these to work in order play a game, there is merit in looking at CPU limited games (ie UT). Like I said CPU limited bench marks are not ideal as the cards raw horse power gets washed into a bunch of other things, but it is the interaction of the vid, drivers, ect that makes these games work. And if there is a difference be it an API or driver, then that should be noted? Obvisoly you can't bench mark all CPU limited games. My point is 3d Mark and fill rate limted games give us a warm fuzzy for what the card can do, but as always in other enviroments, the over all preformance may not be the same in CPU limited games. I am tired of seeing people bass their decsions off the 3DMark or Q3 scores. Check out the forum on Beyond3d with an open mind is all I hope you will do... ">How did you come up with the general >stability & Ease of Use? Gen. Stab - The Voodoo5 5500 crashed on some occasions. Ease of use - The voltage connector is simply a drag >Then you knock them for the power connector. " Ok look at this link and you will see how many issues there are with the GF2. <A HREF="http://www.insanehardware.com/articles/nvidiacompatible.htm" TARGET=_blank>http://www.insanehardware.com/articles/nvidiacompatible.htm</A> I am not saying that this is a bad thing. But the V5 dose work with these so I would recommending a increase to the v5 compatibility. Putting the power connector has helped them to Aviod these power problems from the APG buss (not the power supply). I think that was smart of them. In fact I remember seeing power connectors on some of the first GeForce boards :-) "The fact however remains that here in the Netherlands we have to pay 400 bucks for a video card which is simply not worth that kind of money."Ouch, 400 is a lot of money in any country! " One thing that disturbes me though is that you call this review 'not objective' ... while the article in all waysis "I came to this as your started out with a joke (ok maybe it was not meant to be a joke, but that is what I thought). That is not so bad. If that is your style, then cool. But in at least 4 places you downplay FSAA as saying "In my opinion..." Again nothing wrong with this. But when you repeatedly say "In my opinion.." The review now becomes biased in your opinion. Your opinion is a valid one but keep it in the forums or at the end of the review in a side note. This way the review is now objective. " JB, thanks for your comments, I really do appreciate constructive feedback. I stand behind my review and will back it up in any way I can. However, I can not agree with a lot of what you are saying, but that's ok!, those are your findings/opinions and what is noted in the review are mine."No thank you for taking the time to do the review. You know your stuff and have a good grip on what is going on. I am not trying to change your opinion or fight a Holy war for 3DFX. I just want to see thing as they are for all of the good parts and bad parts. So many reviews DO NOT DO THAT! And that is sad. I honestly felt the scores are a bit low in some of areas or not constant to other reviews. But it is your review and that is the way you saw it.

08-10-2000, 13:02 | posts: n/a

I agree that FSAA in 32bit isen't playable, but what about UT & NFS PU (Both glide 16bit) and playable in 800x600 with 4xFSAA.What about all the old games wich happen to be VERY playable with FSAA and limited to a res of 640x480?Oh and another thing, you will not se a big increase in perfomance with 700mhz thunderbird if you play at 1024x768 32bit with 4xFSAA...A last note: I prefer lower resolution with FSAA than a higher res without... What about those with 15" monitor who can't play at very high resolutions?

01-18-2001, 22:46 | posts: n/a

If you want to get rid of that V5, I'll gladly take it off your hands. No. Seriously. I am a 3dfx fan, and can't find one anywhere. I have a lot of stuff to trade, so E-mail me @ tigerboy15x@netzero.net Thanks for a great site!ThRaShEr19

02-26-2001, 05:23 | posts: n/a

Jb i tottaly agree with you!I used to have a voodoo 3000 till few months when i upgrated it.When the geforce 2 gts was released everyone seemed to say that it kicked ass and that voodoo 5500 sucks!I was impressed i i was a 3dfx man till then.So i upgrated my voodoo with this supposed kick ass gf 2gts.Now as soon as i installed it 1/3 of my games would not work!Why...invalid video mode msges,would not work at all,low frame rates...I was so dissapointed...I searched for patches to support my Creative gf2 but nothing.ie falcon 4.0 would not work,superbike 2001 and gp 3 would give me crap fps and colours.I made couple of formats...newest drivers but still nothing.This is when i took my supposed "Super" gf back into the shop and changed it for the voodoo 5 5500.Ahhhh at last everything seemed normal again!all games would work and especially flight sim 2000 which worked better now than with the gf.I hate nvidia for the trouble it got me to!Since 3dfx is closed i will only buy again from nvidia when 3dfxs technology will be used in future products of nvidia...