Fermi's lead over Cypress shrinks w/ new drivers

[v8envy]

According to:

http://www.techspot.com/review/283-geforce-gtx-400-vs-radeon-hd-5800/

Tested with the 257 vs 10.5 drivers. Fairly small selection of new and one or two year old titles. Using FRAPS to measure frames during "real" gameplay vs. results reported by timedemos.

The takeaway message is: for real gameplay the competing cards are completely indistinguishable. In the two cases where Fermi has a clear advantage the frame rates are too low to be playable in the first place, so at playable settings there is no practical difference.

Which means the difference in real world gameplay is more a factor of how big of an OC you get on your particular card.

Now, how could this happen? There a few theories:

1. NV has recalled all review sample cards and this testing was done with retail hardware. Retail card BIOSes may have been tweaked (read: made slower) so that retail cards are possible in the first place.

2. 257 driver is a showcase for optimization tradeoffs. In order to make a few commonly reviewed titles faster optimizations were made which decrease performance in other cases.

3. Testing fail

 

[SlowSpyder]

Interesting read. I wonder if they'll be 'blacklisted'.

When you consider the higher price, heat, noise, power draw all for a similar experience it seems that the 58xx cards are really a bargain compared to Nvidia's parts. You could make an arguement for the 480 I guess, it is generally faster (though why does this $100 more expensive part lose so often to the 5870 in their tests?) but the 470 was a huge let down compared to the 5850.

 

[happy medium]

According to:

http://www.techspot.com/review/283-geforce-gtx-400-vs-radeon-hd-5800/

Tested with the 257 vs 10.5 drivers. Fairly small selection of new and one or two year old titles. Using FRAPS to measure frames during "real" gameplay vs. results reported by timedemos.

The takeaway message is: for real gameplay the competing cards are completely indistinguishable. In the two cases where Fermi has a clear advantage the frame rates are too low to be playable in the first place, so at playable settings there is no practical difference.

Which means the difference in real world gameplay is more a factor of how big of an OC you get on your particular card.

Now, how could this happen? There a few theories:

1. NV has recalled all review sample cards and this testing was done with retail hardware. Retail card BIOSes may have been tweaked (read: made slower) so that retail cards are possible in the first place.

2. 257 driver is a showcase for optimization tradeoffs. In order to make a few commonly reviewed titles faster optimizations were made which decrease performance in other cases.

3. Testing fail

Funny every other site that did testing with the gtx465 and the BETA 257's showed a major increase with the other cards.
This must be some PR stunt from ATI since there 10.4 and 10.5 drivers gave NO performance increases.
This just shows the nvidia drivers must be better then we thought.
Wonder why Anandtech is not re-reviewing?
This website= fail

 

[KIAman]

Don't worry, nVidia will soon RELEASE THE KRAKEN which will consist of several Fermi GPUs and pwn AMD face for the performance crown.

 

[SlowSpyder]

Funny every other site that did testing with the gtx465 and the BETA 257's showed a major increase with the other cards.
This must be some PR stunt from ATI since there 10.4 and 10.5 drivers gave NO performance increases.
This just shows the nvidia drivers must be better then we thought.
Wonder why Anandtech is not re-reviewing?
This website= fail

If I didn't know better I would swear you were Wreckage. :O


Wait, I don't know better...

One thing worth mentioning about their tests is that they did a manual walkthrough in games, so this is real world results so to speak... but, at the same time it would be difficult to reproduce the exact same movements in a manual walkthrough. So, this isn't exactly scientific. Though I'm sure these results are 'in the ballpark'.

 

[happy medium]

Interesting read. I wonder if they'll be 'blacklisted'.

When you consider the higher price, heat, noise, power draw all for a similar experience it seems that the 58xx cards are really a bargain compared to Nvidia's parts. You could make an arguement for the 480 I guess, it is generally faster (though why does this $100 more expensive part lose so often to the 5870 in their tests?) but the 470 was a huge let down compared to the 5850.

Funny I remember the gtx 470 being ....what did you guys say? 5% faster then a 5850.

Latest test show it just under or equal to a 5870 now.

http://www.anandtech.com/show/3745/nvidias-geforce-gtx-465/4

Really you guys sound like a add campaign for ATI.....even worse then me.


Edit : what do I have to list all the other sites that show an increase?
I know, I know the are all Nvidia fanboy sites. You are really grasping at straws here.

I said my peace, now go on with your advertisement and lie to everyone.
Good day.

 

[SlowSpyder]

Like I said, I think the difference is possibly due to the built in tests, where AMD/Nvidia might be able to optimize somewhat vs. real world gameplay.

I'm just posting about what was found from this website, no need to get so defensive and angry sounding... I'm not here to 'lie to everyone' or call other sites 'Nvidia fanboy sites'. This thread is about what this website found using a different method of testing (again, real actual gaming). Take from it what you want...

 

[Lonyo]

Funny every other site that did testing with the gtx465 and the BETA 257's showed a major increase with the other cards.
This must be some PR stunt from ATI since there 10.4 and 10.5 drivers gave NO performance increases.
This just shows the nvidia drivers must be better then we thought.
Wonder why Anandtech is not re-reviewing?
This website= fail

So the fact that their own testing with different demos (in the same games) had showed a 16% difference means the site is fail?

http://www.techspot.com/review/283-geforce-gtx-400-vs-radeon-hd-5800/ Site is fail.

http://www.techspot.com/review/263-nvidia-geforce-gtx-480/ Site is not fail.

Given that (albeit with different testing) BFG10k (who must also be fail) has claimed that NV seem to focus on popular titles, them focusing on timedemos of popular titles doesn't seem like a massive stretch.
They even say they are surprised at the results. Which is also fail.

 

[railven]

Like I said, I think the difference is possibly due to the built in tests, where AMD/Nvidia might be able to optimize somewhat vs. real world gameplay.

I'm just posting about what was found from this website, no need to get so defensive and angry sounding... I'm not here to 'lie to everyone' or call other sites 'Nvidia fanboy sites'. This thread is about what this website found using a different method of testing (again, real actual gaming). Take from it what you want...

Makes you wonder why he would get upset?

Being all real life performance numbers, this is a good bench to give you a rough idea of what may happen.

Reminds me of Toyota's thread (aside the CPU bottlenecks) he assumed better probably based on synthetic tests. This article reflects his results and overall disappointment in some of his tests.

Interesting find.

 

[happy medium]

I'm not angry look at your post..
count the negatives that have nothing to do with drivers.

blacklisted
heat ,higher price , power draw,
5850 series a bargain compared to nvidia
100$ more exspensive part thats slower then a 5870 (you gotta be kidding)
gtx 470 a HUGE letdown

If that is'nt a TROLL, biased post with nothing to do with the topic I don't know what is!

Facts: gtx 480 is now 20/25% faster overall then a 5870 and cost 20% more
The price performance ratio of a gtx 470 and 5850 is right on the money, in fact its slowly catching a 400$ 5870.
Heat is the problem if you don't have good air flow in your case.
One other problem is they might not play old games from 3/5 years ago over 100fps but play direct x 11 game great.

I want nothing to do with this thread see ya!

 

[Meghan54]

I'm not angry look at your post..
count the negatives that have nothing to do with drivers.

blacklisted
heat ,higher price , power draw,
5850 series a bargain compared to nvidia
100$ more exspensive part thats slower then a 5870 (you gotta be kidding)
gtx 470 a HUGE letdown

If that is'nt a TROLL, biased post with nothing to do with the topic I don't know what is!

Look in a mirror....you'll see one.

 

[SlowSpyder]

I'm not angry look at your post..
count the negatives that have nothing to do with drivers.

blacklisted
heat ,higher price , power draw,
5850 series a bargain compared to nvidia
100$ more exspensive part thats slower then a 5870 (you gotta be kidding)
gtx 470 a HUGE letdown

If that is'nt a TROLL, biased post with nothing to do with the topic I don't know what is!

The 'blacklisted' comment was a joke, hence the smiley (that didn't show up correctly, but is still there, so I = fail?).

The extra heat, higher price, and extra power draw (you forgot extra noise!) are all negatives. Do you view those as positives?

Going by THOSE benches, the benches the topic of this thread is about, the 58xx cards do look like a bargain. The only tests that I would say the 480 was clearly better was the Metro 2033 benches.

Again, going by the link in the OP, how could you say the 470 is any better than the 5850? Going by their unscientific but real world gaming tests the 5850 is as fast to faster, cheaper, quieter, users less power, and puts out less heat. Wouldn't you consider a card that comes out ~5 months later with those issues a pretty damn big let down?

 

[Lonyo]

Funny I remember the gtx 470 being ....what did you guys say? 5% faster then a 5850.

Latest test show it just under or equal to a 5870 now.

http://www.anandtech.com/show/3745/nvidias-geforce-gtx-465/4

Really you guys sound like a add campaign for ATI.....even worse then me.


Edit : what do I have to list all the other sites that show an increase?
I know, I know the are all Nvidia fanboy sites. You are really grasping at straws here.

I said my peace, now go on with your advertisement and lie to everyone.
Good day.

Honestly, I think you should list them all.
And then look at the settings being used.
Here is Xbitlabs (a fail website, by your reckoning), and their review with Crysis benchmarks at 1920x1200 with 4xAA (same as the site in the OPs link), although this time with 16xAF
http://www.xbitlabs.com/articles/video/display/geforce-gtx-465_7.html#sect4

You might notice (might, or you could ignore it and call it fail) that the GTX470 underperforms the HD5850, which is the same as in the OPs review.
Both reviews use 257.15 and 10.5's for the respective cards. They are both different sites.
And both disagree with the AT results, the AT results which seem to lack AA.

No, either the AT scores are the anomaly, or both Xbitlabs and the site in the OP are both wrong (while using the same settings), or the fact that sites use different settings contributes to varying scores and varying relative performance.

Now, you say you could list all the sites which show the site in the OP to be wrong, so I think maybe you should in fact do that, and look at the settings etc being used by the sites.
I found on the first site I looked at results which agree with the site in the OP, and disagree with the AT results. If you want to say Xbitlabs is fail, then feel free, but then you are the one being a fanboy and picking and choosing the sites which show NV in a favourable light rather than actually looking at the benchmarks and their implications (like different cards are better with different settings).

Or you could just say every site which doesn't paint NV with a rosy glow is fail.

 

[Unkle_Tar]

Hard[OCP] showed at launch that there was very little difference in the cards when it came to real world gaming, not canned bechmarks. This just reinforces that conclusion.

 

[brybir]

Happy, I honestly don't know what to make of you. Ive lurked for a long time and I know that every time a post is made about Nvidia having issues you will be one of the first few posts somehow saying exactly the opposite of the linked article and how it is actually a "win" for Nvidia. You remind me of wreckage in that sense.

This article was done with FRAPS and with human players in actual gaming scenarios. The results are what they are and we should all take them for what they are worth, whatever that may be. But to simply dismiss them as being "incorrect" and then immediately calling this an ATI "stunt" while implying that we/OP/article should continue in our group delusions and lies makes you just seem childish and confrontational when their was no intent in this thread to create such an atmosphere.

So, with that said, what would your post sound like if it was not written like you are getting paid by Nvidia? Lets try that starting with your post:

Funny I remember the gtx 470 being ....what did you guys say? 5% faster then a 5850.

Latest test show it just under or equal to a 5870 now.

http://www.anandtech.com/show/3745/nvidias-geforce-gtx-465/4

Really you guys sound like a add campaign for ATI.....even worse then me.


Edit : what do I have to list all the other sites that show an increase?
I know, I know the are all Nvidia fanboy sites. You are really grasping at straws here.

I said my peace, now go on with your advertisement and lie to everyone.
Good day.

Now, lets try to revise that:

This is an interesting article to say the least. I was under the impression that the 470 and 480 were showing significant gains over the competing 5xxx series from ATI with the release of the new Nvidia drivers while the new ATI drivers did not indicate performance increases.

From reading other review sites over the past month it was my understanding that right now, on average, the 470 is 5% faster than a 5850 and more recently, near or equal to the 5870. I think these driver changes have done a great job of closing the gap between the cards and I am truly excited for the competition between vendors so that I can enjoy excellent performance at reasonable prices!

Sometimes I do wonder though whether some of these pages like the OP linked are just ATI PR stunts, I mean, why is it necessary to do a weird time demo with FRAPS being that both ATI and Nvidia can optimize their drivers for demos. Maybe we should assume that they both do that anyways?

This article almost made me want to accuse some of you of being ATI fanboys and perpetrating an ad campaign against Nvidia's products but I know that would be ridiculous. Sure, Nvidia and ATI have done similar things in the past, but I respect each and every one of your intelligence enough to know that the community would eventually see right through that, especially given the oddly polarized nature of the graphics cards forums. Besides, for the life of me, I cannot figure out why anyone would take sides for a graphics card vendor in the way many cheer for their local sports teams.

But, I have written a lot now, and I should be finishing up my work at my day job which is not at Nvidia or ATI PR, so I can get home to my wife and kids and have a meaningful evening enjoying the outdoors and some PC video games.

 

[GaiaHunter]

Happy, I honestly don't know what to make of you. Ive lurked for a long time and I know that every time a post is made about Nvidia having issues you will be one of the first few posts somehow saying exactly the opposite of the linked article and how it is actually a "win" for Nvidia. You remind me of wreckage in that sense.

This article was done with FRAPS and with human players in actual gaming scenarios. The results are what they are and we should all take them for what they are worth, whatever that may be. But to simply dismiss them as being "incorrect" and then immediately calling this an ATI "stunt" while implying that we/OP/article should continue in our group delusions and lies makes you just seem childish and confrontational when their was no intent in this thread to create such an atmosphere.

So, with that said, what would your post sound like if it was not written like you are getting paid by Nvidia? Lets try that starting with your post:



Now, lets try to revise that:

Don't go by that road.

Not that I don't agree with you but this kind of stuff will only be seen as personal attacks and get mod flak.

And if you need to appeal to someone to use common sense you are losing your time - either they don't have or if they have they are choosing not to use it.

 

[happy medium]

Honestly, I think you should list them all.
And then look at the settings being used.
Here is Xbitlabs (a fail website, by your reckoning), and their review with Crysis benchmarks at 1920x1200 with 4xAA (same as the site in the OPs link), although this time with 16xAF
http://www.xbitlabs.com/articles/video/display/geforce-gtx-465_7.html#sect4

You might notice (might, or you could ignore it and call it fail) that the GTX470 underperforms the HD5850, which is the same as in the OPs review.
Both reviews use 257.15 and 10.5's for the respective cards. They are both different sites.
And both disagree with the AT results, the AT results which seem to lack AA.

No, either the AT scores are the anomaly, or both Xbitlabs and the site in the OP are both wrong (while using the same settings), or the fact that sites use different settings contributes to varying scores and varying relative performance.

Now, you say you could list all the sites which show the site in the OP to be wrong, so I think maybe you should in fact do that, and look at the settings etc being used by the sites.
I found on the first site I looked at results which agree with the site in the OP, and disagree with the AT results. If you want to say Xbitlabs is fail, then feel free, but then you are the one being a fanboy and picking and choosing the sites which show NV in a favourable light rather than actually looking at the benchmarks and their implications (like different cards are better with different settings).

Or you could just say every site which doesn't paint NV with a rosy glow is fail.

xbit is ok, Look at the rest of the benches, the gtx 470 is beating the 5850 by more then 5% OVERALL. Much more.
I'm not talking games that are allready old games like resident evil where it really dosent matter. There over 100fps anyway.
ANd metro 2033 they failed to use aa because thats where the lead will increase.

 

[happy medium]

Happy, I honestly don't know what to make of you. Ive lurked for a long time and I know that every time a post is made about Nvidia having issues you will be one of the first few posts somehow saying exactly the opposite of the linked article and how it is actually a "win" for Nvidia. You remind me of wreckage in that sense.

This article was done with FRAPS and with human players in actual gaming scenarios. The results are what they are and we should all take them for what they are worth, whatever that may be. But to simply dismiss them as being "incorrect" and then immediately calling this an ATI "stunt" while implying that we/OP/article should continue in our group delusions and lies makes you just seem childish and confrontational when their was no intent in this thread to create such an atmosphere.

So, with that said, what would your post sound like if it was not written like you are getting paid by Nvidia? Lets try that starting with your post:



Now, lets try to revise that:

I see what your saying, and apologize to you.

 

[railven]

I have a question, and maybe people here might have an answer.

These are real game numbers performed on a system where the CPU and other hardware may not or be a small bottleneck.

How do you think these numbers would correlate to say someone with a Phenom II X3, or a X4, or a Core 2Q.

Do you think in such scenarios the 400 series may overcome the 5000 series? Just curious.

I wonder if the results would scale to show similar performance (although granted lower.)

 

[happy medium]

I have a question, and maybe people here might have an answer.

These are real game numbers performed on a system where the CPU and other hardware may not or be a small bottleneck.

How do you think these numbers would correlate to say someone with a Phenom II X3, or a X4, or a Core 2Q.

Do you think in such scenarios the 400 series may overcome the 5000 series? Just curious.

I wonder if the results would scale to show similar performance (although granted lower.)

I think Russian had a thread about cpu scaling and ATI /Nvidia cards and the ATI cards were better with slower cpu's.
Mabe he will chime in and explain.

 

[Lonyo]

xbit is ok, Look at the rest of the benches, the gtx 470 is beating the 5850 by more then 5% OVERALL. Much more.
I'm not talking games that are allready old games like resident evil where it really dosent matter. There over 100fps anyway.
ANd metro 2033 they failed to use aa because thats where the lead will increase.

I addressed the game you linked to from AT, Crysis.

What this does show, more than anything else, is that you need to be really selective when picking a card.
Whether the HD5850 or GTX470 is faster (or better value for money) really does depend on what you play, what resolution you play at, and what specific settings you use.
Saying "on average it's 5%" or "on average it's 20%" is meaningless when there is so much variety between what one site uses as its settings, and what another site which shows different results might use.
It's not anti-NV, it's not ATI PR, it's not anyone trying to make something look bad, it's just the case that more than ever (it seems) there is a lot of variation between how ATI and NV cards perform with a wide variety of different settings and benchmarks, which makes it a lot more complicated to put forward a value proposition using blanked approximate performance differences from websites.

Yes, typically the GTX470 and GTX480 are faster than their less (oops typo) expensive counterparts, and it can be up to 25%, but even in a single game it can vary from being slower to being faster.
Which means that you can put forward support for either side and play the value and performance game until you're blue in the face, and argue both sides equally, but at the end of the day, what matters more for an individual user is not that the GTX480 can be on average 25% faster if you use the right benchmark, or that the HD5850 can be faster than the GTX470 if you use the righ benchmarks, but look at which card performs best in the specific selection of games you are interested in at your specific resolution using the settings you are most likely to use (e.g. AA/AF levels).

That, more than anything, is what this shows (although it makes it in no way truer than it was before, but it's worth repeating anyway).
It's also why most threads where people ask for a card recommendation break down, because everyone has their own personal preference and can show benchmarks putting forward a certain card to be better value.

 

[luv2increase]

The real test would be to pit 2 x 480s up against 2 x 5870s. I don't think 1 x 480 is a smart buy. It is only when 2 are bought where the real benefits begin solely due to the extreme scaling which is given. When you think about it, many people with enough money to buy a 5870 or 480 usually want the best and will purchase a 2nd or 3rd for SLI/CF. This would be the real scenario to test.

I'm not a fanboy of either camp. I am first and foremost an enthusiast. Everyone who knows me knows this. I recently sold 3 x 5870s to test the waters with Nvidia's latest offering. With a single 480, I wasn't too content. Now, with 2 x 480s, I appreciate them as much as I did 3 x 5870s. Plus, I get the benefits of CUDA based GPGPU applications. This was actually the "main" reason I switched to the 480s. It was for Adobe's Mercury Playback Engine in their latest CS5 offering.

So, I beg Anandtech to do a test between 2 x 5870s and 2 x 480s. I will also plead that they do so with Nvidia's next WHQL driver. That way we can have 10.5 whql vs 257.xx whql.

 

[brybir]

I see what your saying, and apologize to you.

Ive noticed that a few people here seem to really pull your strings for whatever reason but I was just making the overall point that we could all benefit from a more civil discussion about these issues. But, if it were entirely civil, I would miss the dramatic moments when the mods come in and block and ban people...its like when the cops show up at the neighbors after they are fighting and the someone goes out in handcuffs....fun to watch but bad in general.

Then again this is the video card forum....as someone in another thread said, the wild west of forums at AT.

 

[busydude]

I have a question, and maybe people here might have an answer.

These are real game numbers performed on a system where the CPU and other hardware may not or be a small bottleneck.

How do you think these numbers would correlate to say someone with a Phenom II X3, or a X4, or a Core 2Q.

Do you think in such scenarios the 400 series may overcome the 5000 series? Just curious.

I wonder if the results would scale to show similar performance (although granted lower.)

This article might interest you. Look at graphs for resolutions at 1080p or greater.

http://www.xbitlabs.com/articles/cpu/display/cpus-and-games-2010_4.html#sect0

As for specific recommendations concerning the choice of a graphics card, the current Radeon HD 5000 series from AMD is overall less CPU-dependent than the GeForce GTX 400 and 200 series from Nvidia. This trend does not hold true in every game, however, and you can see the opposite situation in certain games.

Also read: http://www.tomshardware.com/reviews/balanced-gaming-pc-overclock,2625-8.html (Intel cpu scaling)
http://www.tomshardware.com/reviews/balanced-gaming-pc,2477-8.html (AMD CPU's with previous gen gaming cards)

 

[GaiaHunter]

This article might interest you. Look at graphs for resolutions at 1080p or greater.

http://www.xbitlabs.com/articles/cpu/display/cpus-and-games-2010_4.html#sect0

The Toms series of reviews about building a balanced gaming PC (no GTX400 in there though) also show that the GTX280 and GTX295 seems to depends on more CPU power to reach its full potential.

http://www.tomshardware.com/reviews/balanced-gaming-pc-overclock,2625.html and previous.