V5 Vs geforce2

[lordneo99]

I know i will get at least some flames, but i am
asking ,
why does 3dfx need 2 graphics chips to
(depending on your point of view)
equal/surpase/blow the 1 chip implementation of the geforce2 ?
if the battle was 1 chip to 1 chip , then id say
we'd have a fair comparison

basically that means its more expensivce to implement ,nothing more then beefed up Voodoo1
at chickshardware.com
the V5500 gets beaten up by both the geforce2 and radeaon.
MDk2 the voodoo cant even play.
cause it has NO T&L

http://www.chickshardware.com/html/reviews/radeon/6.html

and slow in quaver, which is quake3 High setting on steriods

http://www.chickshardware.com/html/reviews/radeon/5.html

 

[Aboroth]

What does it matter?
Who cares if it takes more than one chip? If they can give me good performance in my price range I don't care if it takes 3000 chips (as long as they fit in my computer. )
Also remember that the VSA-100 was meant to be scaleable. It is really hard to do that.

 

[han888]

and also the voodoo5 have 64 meg, if we compare with 32meg gts, the gts still the winner

 

[Esis]

The Voodoo5 is beaten soundly by the Radeon and the Geforce2. My Geforce2 (Hercules Prophet II) obviously beats it, and the Radeon beats it right out of the box, even with young drivers. I'm not saying the Voodoo5 is a bad card, but it can't compare to the Radeon or Geforce2.

 

[GlennHoback]

Remember that the V5 was designed to be the 3DFX 'mid-price' entry, along with the 4500 and 6000. A truer comparison would be the GF2/Ultra and Radeon to the 6000 (which, of course, is not yet released). The 5500 was, i believe, meant to compete against the SDR/DDR vid cards, but got released late. This is not trying to excuse 3DFX for being late in their product release cycle, but let's be honest and compare cards which are meant to be comparable - the 4500 to the MX, the 5500 to the original GeForce cards, and the 6000 to the Ultras. Of course the V5 (the current top shelf 3DFX product available to consumers right now) is going to get it's butt kicked by the GF Ultra, because the 6000 isn't yet available to compare. The true question will be how the 6000 will stand up the the Ultra, when and if it ever comes out.

 

[BFG10K]

3dfx needs to throw more CPUs on their boards just to stop slipping off the benchmarking charts. They claim to have solved the memory bandwidth problems by doing so, but they haven't solved anything.

All they have done is introduced the overhead of multi-processing, wasted RAM through duplication and increased the costs/heat/power requirements of their boards.

 

[RoboTECH]

<< why does 3dfx need 2 graphics chips to (depending on your point of view)
equal/surpase/blow the 1 chip implementation of the geforce2 ? if the battle was 1 chip to 1 chip , then id say
we'd have a fair comparison
>>



1) The GTS is an absolutely phenominal *core*, but the memory bandwidth holds it back. An SLI implementation is a major helper for increasing bandwidth. Of course, 3dfx did take an &quot;easy&quot; way out by making the VSA-100's have only 2 pixel pipes without a 2nd TMU (my primary complaint with the card)

2) Why compare 1 chip to 1 chip? They cost about the same.



<< MDk2 the voodoo cant even play cause it has NO T&amp;L >>



Hardware T&amp;l is not a requirement, and the 5500 can play MDK2. It doesn't play it as well, but it still plays it quite well. I've found that MDK2 is not that much of a framerate hog.



<< and slow in quaver, which is quake3 High setting on steriods >>



I checked that link, and his results were about 12 fps slower than what I got in quaver with the 5500 when I had it.



<< I'm not saying the Voodoo5 is a bad card, but it can't compare to the Radeon or Geforce2. >>



well, I've had both a 5500 and a 64MB GTS (as well as a 32MB GTS) and there is far more to life than Q3 scores. That being said, the 5500, with a few driver tweeks, gets outstanding image quality and absolutely flies in Q3.



<< 3dfx needs to throw more CPUs on their boards just to stop slipping off the benchmarking charts. They claim to have solved the memory bandwidth problems by doing so, but they haven't solved anything.

All they have done is introduced the overhead of multi-processing, wasted RAM through duplication and increased the costs/heat/power requirements of their boards. >>



well, for once BFG, I agree with you (mostly)

It had to happen....

 

[Fierysonic]

I know that I'm going to get flamed for this, but comparing the V5 to the Ge2 GTS is like comparing a Celery to a Duron. The GTS and the Duron whip the V5 and the Celeron in every test that you can think of, except for a couple, like FSAA. People just like inferior products and refuse to look at the facts. Just my 2 cents.

 

[RoboTECH]

<< I know that I'm going to get flamed for this, but comparing the V5 to the Ge2 GTS is like comparing a Celery to a Duron. The GTS and the Duron whip the V5 and the Celeron in every test that you can think of, except for a couple, like FSAA. People just like inferior products and refuse to look at the facts. Just my 2 cents. >>



well, let's skip the Duron/Celeron thing, as I totally agree.

But as you can see, I'm a GTS 64MB owner, and after using the 5500 for some time, I can hardly say that the GTS &quot;whips&quot; the 5500.

The 5500 actually runs ALL the games I have played with excellent (in some cases phenominal) visual quality and excellent speed. I noticed very brief slowdowns in MDK2 on very rare occassions, but that's it. It was also completely stable and had outstanding high-res 2d quality.

The GTS has been a card of extremes for me. In some cases (MDK2) it is incredible. Crank the resolution and go. In other cases (NFSU and Deux Ex) it has been no better than a V3 I had. Driver instabilities and poor TC implementation have pretty much negated the speed advantage the GTS had over the 5500 in many situations, and it's high-res 2d desktop quality is rather poor, especially by comparison.

Again, having owned and used both cards for a few weeks, it troubles me to see completely mis(un) informed comments like that one.

Try playing games with the framerate counter turned off. Set things up so that you get equally excellent visual quality for both cards. Now play for awhile. Do you notice slowdowns? If not, then don't sweat the benchmarks.

You might be surprised. There is more to choosing a car than it's 0-60 MPH time (as long as both are 'fast enough')

 

[Mem]

I am not going to start a flame war here ,but can you image if Nvidia &amp; 3DFX teamed up,that would be a force to reckon with(I know this will not happen ,but then again ,you could say the same thing about Aureal being sold to Creative Labs ,not the samething but still a surprise).

 

[Aboroth]

I have no delusions on my Voodoo5. I know that the GTS beats it in a lot of benchmarks. However, I don't play Quake 3. Or UT. Or MDK2. Those are just stupid benchmarks to me. The games I do play get surprisingly great image quality with much more than adequate speed, and I also get superb 2D, which is all I ask for and expect from this card.

I bought a video card that meets my wants. You bought a video card that does what you want. I accept that you are happy with your video card. I am happy with mine. You are happy with yours. Leave it at that and stop attacking everyone else's decisions. These discussions get very childish.

People just like inferior products and refuse to look at the facts.

I am not here to defend the Voodoo5. You can go elsewhere to read pages and pages of crap for it and against it. However, that statement suggests that the people who bought the Voodoo5 are ignorant idiots. That is not a good way to make friends. Maybe you are ignorant for not realizing that there are things the Voodoo5 is more suited for than the Geforce 2. Of course, some things are &quot;better&quot; on the Geforce 2, and hopefully you bought it for a good reason.

Anyway, why is it so important to convince others that your card is so much better than theirs? People who base their whole self worth on what hardware they own should go outside every once in a while or talk to a psychiatrist.

 

[Audiofight]

I currently own a V5 5500. I have been disgusted with the performance since day one. I read all the reviews and heard it talked up. So I went and bought one. Big mistake in my book.

Today, I just received the GeForce2 MX I ordered for a system I am building for a friend. While I am waiting for the rest of the components, i decided to run a few benchmarks on it. The card was set to default settings, no tweaking on it, and I ran 3DMark 2000. 39xx. Not bad. Unfortunately, my V5 only gets 33xx with fastest performance turned on, and it goes well down hill when i turn on FSAA 2x and 4x. My lowpoint was 710 on 3DMark.

Now I have heard that 3DMark is highly Nvidia biased, but that shouldn't make this big of a performance hit when comparing a 32MB GF2 MX and a 64MB V5 5500.

 

[RoboTECH]

Aboroth:


<< People just like inferior products and refuse to look at the facts.
<snip> that statement suggests that the people who bought the Voodoo5 are ignorant idiots....Maybe you are ignorant for not realizing that there are things the Voodoo5 is more suited for than the Geforce 2. Of course, some things are &quot;better&quot; on the Geforce 2, and hopefully you bought it for a good reason. >>


very good point, and why statements like that are ridiculous. I have used both cards extensively, and both cards had some outstanding points and some not-so-outstanding points.


<< People who base their whole self worth on what hardware they own should go outside every once in a while or talk to a psychiatrist. >>


y'know, that's what my shrink keeps telling me. <g>
seriously tho, I like a good debate. I have my football smacktalk group so we can argue about who was the best RB of all time, I have my hockey newsgroup so we can argue about who the best team of all time was, and here I have the Anand-forums, a haven for some very, very self righteous nvidiots (and nvidia fanboys). What's funny is that I spend more time arguing with peeps who have the same hardware that I have, heh....
audiofight:


<< I currently own a V5 5500. I have been disgusted with the performance since day one.... Today, I just received the GeForce2 MX. I ran 3DMark 2000. 39xx. Not bad. Unfortunately, my V5 only gets 33xx with fastest performance turned on, and it goes well down hill when i turn on FSAA 2x and 4x. My lowpoint was 710 on 3DMark. >>


1) What drivers were you using?
2) What tweeks to the drivers did you make?
3) Did you try upping the clock speed?
4) The 5500 performs far better on high-end P3's than mid-range Athlons (Shitty 3dNow implementation)
5) Did you ever figure out how to control that helicopter in 3dMark? I have tried several different joysticks and have YET to figure out how the hell to control that damn thing! And what's with the goofy people walking around in that courtyard near the big ships? I just don't understand how to play that 3dMark2000 game.

 

[Staz]

Since 3dfx likes to put multiple GPU's on their cards, why doesn't Nvidia do the same thing? Could you imagine the speed of a Geforce2 Ultra GTS card with 2 GPU's and 128MB, 64MB for each GPU! If this was possible, that card would rock!!!

Or for a inexpensive version, place 2 Geforce2 MX GPU's on one board along with 64MB, 32MB each.

Thoughts anyone???

 

[RoboTECH]

<<
Since 3dfx likes to put multiple GPU's on their cards, why doesn't Nvidia do the same thing? Could you imagine the speed of a Geforce2 Ultra GTS card with 2 GPU's and 128MB, 64MB for each GPU! If this was possible, that card would rock!!!
>>

Well, it's not quite that easy, for a few reasons:

1) You can't just &quot;slap a couple of GPU's on a card&quot; and call it a dualie. You have to figure out a way and write drivers that will allow the cards to do SLI/AFR/whatever

2) They would have to come up with a unique method of doing this. ATi does Alternate Frame Rendering, 3Dfx does scan-line interleave. unless nVidia wants to suffer the wrath of a lawsuit, they'd have to come up with something new.

3) nVidia has already gone on record as saying that putting 2 GPU's on a card is a horrible idea. It would be some bigtime egg on their face if they were to do this after saying what they've said

4) Their GPU's are huge enough as it is. You think a 5500 is large, just imagine how much logic would be involved in a duel nvidia card

 

[hans007]

Ok.. ask this to yourself. The nvidia card needs very very fast ram, the v5 doesnt. So.... they both have in both ways something that is more expensive to implement

 

[AMB]

I shall start a flame GeForce 2

 

[sandorski]

Regarding the multi-chip issue, the only drawback to that is 3dfx's supply problem. If they had a single chip solution, we'd probably of seen Voodoo4s - 6s by now.

 

[Wingznut]

Hey!!!

&quot;Did you ever figure out how to control that helicopter in 3dMark? I have tried several different joysticks and have YET to figure out how the hell to control that damn thing! And what's with the goofy people walking around in that courtyard near the big ships? I just don't understand how to play that 3dMark2000 game&quot;

(That sure sounds familiar. )

Benchmarks suck. Explain to me just how the number of &quot;3dMarks&quot; can make you enjoy NFS5 more?

Personally... I play games, not benchmarks.

Fierysonic, exactly what game gives you problems with your V5? And why don't you play Deus Ex and UT on both cards, and then give us your opinion.

 

[RoboTECH]

<< Fierysonic, exactly what game gives you problems with your V5? And why don't you play Deus Ex and UT on both cards, and then give us your opinion. >>



he doesn't play games. He just reads benchmarks off a website and then fancies himself an expert, like so many other peeps around here. B-/

 

[Wingznut]

Oops, I meant to direct that question to audiofight.

How about it, audiofight? What games do you have problems with?

 

[BFG10K]

he doesn't play games. He just reads benchmarks off a website and then fancies himself an expert, like so many other peeps around here. B-/

Aw c'mon RoboTECH, you're not that bad!

 

[RoboTECH]

HA!!!

<kicks BFG in shins>

heh...beyotch!