Good Radeon Review

[Taz4158]

Read it here:
Click

 

[snow patrol]

I think the overclocking results seem pretty poor, percentage wise, compared to the improvements I see when overclocking...

 

[Leon]

Nice review, but their benchmarks = BS.

GTS is capable of 60fps + @ 1280 1024 32bit, not 42fps like their graph shows. (My GeForce 256 DDR can do 42fps easily).

Leon

 

[pdo]

oh no it's gonna turn into a flame war now.

 

[BenSkywalker]

"Oftentimes the Radeon has higher framerates in 32bit that the GeForce2 GTS did in 16bit."

So often in fact, it never happened.... not once........ not even close

"This first set of numbers comes from the MadOnion 3DMark 2000. While the resulting numbers are not as impressive as they could be, it's important to note the huge gain in performance you get when you move from 16bit to 32bit mode. Amazing is the first thing that comes to mind. I have yet to see ANY card where performance INCREASED when in 32bit mode over 16bit."

So huge of a performance gain in fact, that it slows down in every 3DMark2K score moving from 16bit to 32bit

The only test that the Radeon is faster in 32bit then it is in 16bit is the Quake3 Fastest test which if you look, the GF2 is also(~20-40FPS gain moving from 16bit to 32bit @640x480, yeah, OK, sure). Very interesting to see that there PIII 1GHZ with GF2 or Radeon is slower then my Duron 650 with GeForce DDR.

 

[BFG10K]

16bit gaming yields NO performance increase (more often a decrease) and the image quality is only average. I see no reason a person would even run this card in anything other than 32bit mode.

Correction: 16bit gaming on the Radeon yields no performance increase and also looks abysmal. I see no reason why anyone in their right mind would run 16 bit mode on a Radeon.

Also the HQ 32 bit scores for the GTS were far too low. Tom with his 32 MB GTS gets:

1024 x 768 x 32: 95 FPS (HardOCP 75.2 FPS).
1280 x 1024 x 32: 54.7 FPS (HardOCP 47.8 FPS).

And that's with the ancient 6.16 drivers.

 

[pidge]

Since Kyle didn't write the article, I place little weight on the articles benchmark numbers. It is far less than I got when i had a Geforce 2 GTS with a 933MHz PIII. I won't cap on the Radeon 32MB DDR because for the price, it is an excellent card. But I am really happy with my Geforce 2 Pro and there is no reason why I would rather have a Radeon over any Geforce 2 GTS, Pro or Ultra. I simply have no problems with the Geforce 2 series of cards. And it is nice to play at 1280x1024 or higher with complete smoothness.

 

[snow patrol]

Yeah, the benchmark results are extremely fishy all round really...

 

[Taz4158]

Funny how all the nVidia people have to crap on a positive Radeon review because the Benches CAN'T be correct. So they must be fraudulent. Let it rest guys. I don't see people in the Radeon camp denouncing positive GTS reviews. I'm quite sure these are the results that the reviewer achieved and he wasn't "paid off" by ATI. It also seems to be the only review using semi new drivers so more stock can be put in it. Now I have no idea if there are faster drivers than the 6.18 for the GTS since that was about the time I saw the light and switched, but there ARE faster Radeon drivers now than the ones he used. Seems right on the mark with regards to DVD, image quality and 32 bit to me. BTW I get higher results with my Radeon than he did just like you nVidia folks.

 

[oldfart]

The Radeon benchmarks scores look to be correct. The 3Dmark and Q3 scores look very similar to what I get. I notice the Q3 tests don't mention what version of Q3 was tested. Benchmark scores are considerably different from 1.17 to 1.27. This could be a reason for the discrepency. The I know people who refuse to upgrade to 1.27 because it drops their Q3 score.

I do agree that many of the statements make no sense. I read through it yesterday and could not understand what he was talking about.

BFG, Since you have an MX, I'm sure 16 bit performance is very important to you. Radeon owners don't need to worry about 16 bit. They never use it.

I normally wouldn't go in anywhere yapping about the "abysmal" performance and "abysmal" 2D of an MX. But if you want to go around crapping on Radeon threads, I can be an idiot as well.

 

[Taz4158]

Don't forget many of us have Quake tweaked and get higher results too. They're running pretty much stock. That accounts for the differences as well. And BFG I don't recall you ever owning a Radeon, am I wrong?

 

[hans007]

wow all this bickering , its almost like the good old nvidia 3dfx wars. Yes i was a 3dfx loyalist due to 2d quality reasons, and now i've got a radeon. And well, first off hardocp isn't exactlythe most tech savvy place on the net. also the other thing i figured, is if your card can run 32bit color at a playable speed why would ANYONE EVER run 16bit color . please dont bash me, as i have owned a card or cards of every chipset (i'm not counting differing clock speeds like tnt2 and ultra, or geforce2 and ultra) except the geforce 2 mx and tnt1 at some time or another that ATI, 3dfx and nvidia make.

 

[oldfart]

I have owned them all as well. I've owned Rendition, Nvidia, 3dfx, and now my first ATI. I was happy with all of them, and I never went around bashing any of them. Since 3dfx went under, I notice a distant increase of anti-ATI posts. Why the "if its not Nvidia it MUST suck" people feel the need to do this is beyond me. Unfortunately this behavior promotes like responses.

 

[pidge]

I don't think it is a "if its not NVIDIA it must suck" attitude. Alot of us are just not happy (or happy enough) with the Radeon and since the only other performance video card out there is the Geforce 2 series, you get that ATI vs. NVIDIA arguements coming up alot. I never attack Matrox or 3dfx personally but I guess there are others out there who do.

I think the Radeon 32MB DDR is the best card out there for people on a budget but I would rather spend a few extra bucks for the stability and performance the Geforce 2 GTS and higher have to offer over the Radeon. NVIDIA does not have a card which is a perfect card. That is why they offer so many different models. But they make damn good video cards. Why do you think Carmack continues using NVIDIA Geforce 2 cards in his development systems?

Carmack: "NV20 rocks. "

Looks like he is happy with the NV20 too which gives me something to look forward to. ATI makes good Radeon cards too but they lack a few critical things to make me want to go back to Radeon. I am done with Radeon. Maybe when the Radeon 2 comes out, I will give it a shot.

 

[Taz4158]

<< I think the Radeon 32MB DDR is the best card out there for people on a budget but I would rather spend a few extra bucks for the stability and performance the Geforce 2 GTS and higher have to offer over the Radeon >>


Well I and many others I know switched FROM the GTS to the Radeon and have better stability and performance. Guess the odd time it just doesn't work out for some people.

 

[PeAK]

Tom Hardware has always (it seems) had a soft spot for both Matrox and Nvidia. In the following charts and benchmarks, the budget Radeon Products (SDR) clearly beat out the budget MX products soundly in a much &quot;hyped about&quot; Mercedes&quot; benchmark (made in Germany?) in both 16 bit and 32 bit. The most interesting thing about the review is the clear 'restraint&quot; from product bashing often seen in reviews from this site when a popular brand fails to meet the reviews typical &quot;king of the hill is framerate&quot; mark. I know the &quot;tone&quot; of the review if ATI came out near the bottom of the heap and the favoured sons on top. I've been there.

What I am finding today is that people place more importance on word of mouth, forums and direct experience with products rather that trust popular hardware sites of yore implicity. Staff changes/churn mean that the sites, today, are only as good as the open minds of the individuals who run them. Even divesting oneself of the advertising operations using outside companies does not make a site necessarily more neutral or unbiased...it might help, though.

 

[Taz4158]

<< What I am finding today is that people place more importance on word of mouth, forums and direct experience with products rather that trust popular hardware sites of yore implicity. Staff changes and sites are only as good as the open minds of the individuals who run them. >>


It's true. If people can keep their emotions out of it there can be a very healthy sharing of information.

 

[Hawk]

Hey pidge, I know you've had, well, still have, a Radeon, but I know most of the ATI bashers have never had a Radeon, your state:

<<Alot of us are just not happy (or happy enough) with the Radeon and since the only other performance video card out there is the Geforce 2 series>>

probably doesn't apply to lots of people.

 

[silver24]

I to went from a GTS Pro to a 64 Radeon, and the Radeon is a great card and im glad i switched, i see nothing wrong with the benchmarks at all. The Radeon is not that much slower that my Pro was. I do see much better scores when i overclock my Radeon (210-210)than it shows.

 

[pidge]

Hawk,
I still say that &quot;a lot&quot; can apply but I do think the regular ATI bashers are people who have never owned a Radeon. But in long threads, you always see different people saying that they weren't happy with the Radeon because of this or that and that they switched to Voodoo 5 5500 or Geforce 2 and they are happy now. A lot can be relative I guess. I won't really argue to much on that. I like your posts Hawk. I never see your posts as an attack against others or their video cards. Your posts are informative and based more on facts than emotions.



<< It's true. If people can keep their emotions out of it there can be a very healthy sharing of information. >>



Taz,
I think you tend to let your emotions get the best of you to much. You attack to many people who do not agree with you about the quality of the ATI Radeon or who prefer their Voodoo or Geforce cards over the Radeon. I think you should be more like SleepyTim or Hawk. Them two are really great at stating their point without having to flame others.

 

[BFG10K]

Oldfart:

Radeon owners don't need to worry about 16 bit. They never use it.

I never said they did. What I was saying was that the way the article was written, it sounded like the guy was talking about both boards. The GTS sees a huge boost in 16 bit mode.

normally wouldn't go in anywhere yapping about the &quot;abysmal&quot; performance and &quot;abysmal&quot; 2D of an MX.

Because the MX doesn't have have abysmal performance and not all boards suffer from 2D image quality problems. On the other hand all Radeons look crap in 16 bit mode (in fact all of the ATi cards I have seen suffer from this problem).

But if you want to go around crapping on Radeon threads, I can be an idiot as well.

Nobody's crapping anywhere. It's just a simple discussion.

BTW do you post at Rage3D? I ask because there's a guy there called Oldfart but it doesn't sound like it's you.

 

[Leon]

6.18 = Detonator3. Even with this old set, GTS easily do 56fps @1280 1024 bit


Link


That's why HOCP benchmarks is waaaay low. I remeber their V5 review, where they lowered GTS scores....I see a pattern there...

Hard OCP is great website for overclocking...not benchmarking.

Leon

 

[BFG10K]

Taz:

Let it rest guys. I don't see people in the Radeon camp denouncing positive GTS reviews.

That's because the positive GTS reviews don't have incorrect Radeon results. And why should I give it a rest? If I found a similar benchmark with low Radeon scores would you give it a rest?

And BFG I don't recall you ever owning a Radeon, am I wrong?

And? In what way does that affect my comments about low GTS scores and 16 bit image quality/performance on a Radeon?

PeAK:

Tom Hardware has always (it seems) had a soft spot for both Matrox and Nvidia.

A nice excuse but unfortunately Anandtech is getting almost exactly the same scores as Tom.

So, an obviously fraudulent set of GTS benchmarks has been revealed and somehow the ATi users expect me to &quot;give it a rest&quot;. Not only that, but then they accuse me of being an nVidiot for simply commenting on this fact and for commenting about the 16 bit issue with ATi's cards.

Yes, there certainly are zealots in every camp.

 

[ChannelX]

I think that basing your buying decision on benchmarks, unless you are a really hardcore gamer, is not the most intelligent thing to do. I've never had an nvidia board. I replaced my v3-2k with one of the newegg Radeon LE boards and it kicks ass. Smooth as silk and I haven't had any problems.

Based on reviews from owners of both boards the Radeon is a card that is better overall. Might not be the best in all areas but certainly seems so in most....great 3d performance....great video quality....great DVD ability...etc.

The *only* cards that have looked good in 16bit mode are the Voodoo cards. 16bit on the Radeon does suck....but who needs it.

As to the OCP review the guy states that hes an NVIDIA user (geforce2 ultra in his machine) so I hardly think there is a Radeon bias there.

 

[TheTrueEqualizer]

Deleting my post. Forget it. Would cause more arguments.