Poll: GT300 VS. HD5870

[RussianSensation]

Originally posted by: Shaq
Also the FS benchmarks were all at 2560 with 4AA in some cases. It would make more sense to bench at 1680 and 1920 and average the results because the vast majority of gamers have a 22" or 24" monitor. Just showing the highest resolutions and high levels of AA is trying to skew the results in favor of ATI. It is established that the 5870 is better with high AA and high resolutions. But does it really matter if a 285 gets 60 FPS and the 5870 gets 75 FPS in a certain game? They are both very playable. It comes down to cost and availability at that point and for some people power consumption and heat. The 285 can be found for $296 versus the $379 5870 if you can find it. That is 25% cheaper and about how much faster the 5870 is compared to the 285 at common resolutions.

@ Shaq
If you are playing at 1680 or for most games even at 1920, you don't need a 5870 over GTX 285/275/4890 etc.

Right now the 4 most demanding games are Crysis, Crysis: Warhead, and STALKER: Clear Sky and NWN2. You can check out the benches for 3 of these games at 1920 and you'll see that 5870 cleaned up the 285 card. And who buys a $400 graphics card for 1680?

@ MODEL3
Testing at 8AA is not biased. It highlights NVs inefficiency with current architecture in that mode. If you spend $400 on a graphics card, and you can use 8AA with less penalty and get playable framerates, then do so. That is why you paid $400 for a card, not $200.

 

[RussianSensation]

Originally posted by: dguy6789
The limitation on the 5870 is most certainly memory bandwidth.

Don't forget that both NV and ATI usually squeeze another 10-20% through driver improvements over time.

 

[RussianSensation]

Originally posted by: MODEL3

It would be nice if sites had all the meaningful combinations.
Sadly many times this is not possible (for example time constrains)
Especially at launch of a new product, this is very difficult.
I guess, what it would be nice is at least to have some follow up articles.

Check out Xbitlabs 5870 Review - 1280x1024, 1680x1050, 1920x1200, 2560x1600 and minimum frames

 

[MODEL3]

Originally posted by: RussianSensation
@ MODEL3
Testing at 8AA is not biased. It highlights NVs inefficiency with current architecture in that mode. If you spend $400 on a graphics card, and you can use 8AA with less penalty and get playable framerates, then do so. That is why you paid $400 for a card, not $200.

If you do 4AA testing also, i completely agree with what you say.

If you only do 8X AA, it highlights NVs inefficiency with current architecture in that mode, but it does not highlight how much lower is the difference with 4X AA.

The consumer must be informed for all the cases, in order to make an educated decision about what he likes to buy.

 

[MODEL3]

Originally posted by: RussianSensation

Originally posted by: MODEL3

It would be nice if sites had all the meaningful combinations.
Sadly many times this is not possible (for example time constrains)
Especially at launch of a new product, this is very difficult.
I guess, what it would be nice is at least to have some follow up articles.

Check out Xbitlabs 5870 Review - 1280x1024, 1680x1050, 1920x1200, 2560x1600 and minimum frames

Yes, i saw Xbitlabs review.
Although it doesn't cover all the AA combinations (it usually has the most meaningful 4X AA, i like especially the range of the games that tests), Xbitlabs is one of my favorite sites.

We were talking in general what a reviewer in tech site must do in order to inform the public better.

 

[dguy6789]

Originally posted by: Idontcare

Originally posted by: dguy6789
The limitation on the 5870 is most certainly memory bandwidth. Memory bandwidth is the only category in which the 5870 does not outclass the 4870x2 in.

Radeon HD 5870:

Pixel Fill Rate: 27200 MPixels/sec
Texture Fill Rate: 68000 MTexels/sec
Flops: 2720 GFLOPS
Memory Bandwidth: 153.6GB/sec

Radeon HD 4870x2:

Pixel Fill Rate: 24000 MPixels/sec
Texture Fill Rate: 60000 MTexels/sec
Flops: 2400 GFLOPS
Memory Bandwidth: 230.4GB/sec

I would think this theory would be fairly easy to test and either prove or disprove by anyone who has the 5870 in hand.

Just underclock the memory. Do stock, underlocked mem by 10-15%, underclocked mem by 20-30%, and plot the benchmark data.

If you get a linear line with nice R^2 and the slope is nice and large (slope of 1 on a % bandwidth delta versus % fps delta would be perfect) then you can easily conclude the performance would be improved by having higher bandwidth above stock.

What is interesting is that no one seems to have done this test in their reviews yet (or did I miss it?).

What would be even more interesting is if the data proved out this theory as surely AMD would have tested this and determined whether or not bandwidth was really going to be the limitation of the 5870 well before the launch. So why would they artificially handicap the 5870 by knowingly starving it of bandwidth? I don't think we should assume ignorance on their part here.

Indeed.

I am betting that for games where the 5870 is as fast as the 4870x2, memory bandwidth isn't very important. Changing the memory clocks on the cards won't change performance very much. For games where the 4870x2 pulls ahead of the 5870, increasing or decreasing the memory clocks on the 5870 will dramatically effect performance, almost linearly I would wager.

It makes sense though. The 4870 has similar GPU power to the 4850, but 80%~ more memory bandwidth. The 4870 is usually around 20-30% faster than the 4850 as a result, so clearly memory bandwidth is helpful. The 5870 has more than twice the GPU power of the 4870, it actually has twice the GPU power of the 4890. Memory bandwidth only increased from 124GB/sec(4890) to 153GB/sec(5870), 23%~ while GPU power went up by 100%.

 

[bryanW1995]

Originally posted by: Wreckage

Originally posted by: RussianSensation

> 18%[/b] check out Firingsquad for example.

Once again...... the graph i linked used several sites and games. Certainly you could cherrypick a site or game. But I would assume it would be more prudent to look beyond just 1 source. Having read through the 5870 thread and other threads on multiple forums, it looks like around 20% seems fairly accurate.

well, if 5870 is only 20% faster than gtx 285 and is 10% slower than gtx 295, are you saying that gtx 285 is only 30% slower than gtx 295?

 

[bryanW1995]

Originally posted by: Shaq
Also the FS benchmarks were all at 2560 with 4AA in some cases. It would make more sense to bench at 1680 and 1920 and average the results because the vast majority of gamers have a 22" or 24" monitor. Just showing the highest resolutions and high levels of AA is trying to skew the results in favor of ATI. It is established that the 5870 is better with high AA and high resolutions. But does it really matter if a 285 gets 60 FPS and the 5870 gets 75 FPS in a certain game? They are both very playable. It comes down to cost and availability at that point and for some people power consumption and heat. The 285 can be found for $296 versus the $379 5870 if you can find it. That is 25% cheaper and about how much faster the 5870 is compared to the 285 at common resolutions.

you are exactly right about the resolutions. However, if you game at 16x10 the 5870 (and gtx 285) are way overkill today. 60 or 75 fps is unnoticeable to most of us. However, at 25x16 w/4xAA, you could be looking at 25 fps vs 34 fps, a HUGE difference. As future games come out and your video card ages, you will see more and more games that challenge these high end cards at lower resolutions, just like what happened to 8800gtx. So if you game at 16x10 or 19x12 today but you want a card that will last you several years, you can be confident that over time that 30-40% + number will be more accurate at more resolutions/AA levels, plus you'll be able to us DX 11 effects, plus eyefinity, plus extremely low power consumption, etc etc.

The real card to compare with gtx 285 is 5850. That card will have all the same features of 5870 but be priced less than gtx 285 by quite a bit.

 

[bryanW1995]

Originally posted by: Idontcare

Originally posted by: dguy6789
The limitation on the 5870 is most certainly memory bandwidth. Memory bandwidth is the only category in which the 5870 does not outclass the 4870x2 in.

Radeon HD 5870:

Pixel Fill Rate: 27200 MPixels/sec
Texture Fill Rate: 68000 MTexels/sec
Flops: 2720 GFLOPS
Memory Bandwidth: 153.6GB/sec

Radeon HD 4870x2:

Pixel Fill Rate: 24000 MPixels/sec
Texture Fill Rate: 60000 MTexels/sec
Flops: 2400 GFLOPS
Memory Bandwidth: 230.4GB/sec

I would think this theory would be fairly easy to test and either prove or disprove by anyone who has the 5870 in hand.

Just underclock the memory. Do stock, underlocked mem by 10-15%, underclocked mem by 20-30%, and plot the benchmark data.

If you get a linear line with nice R^2 and the slope is nice and large (slope of 1 on a % bandwidth delta versus % fps delta would be perfect) then you can easily conclude the performance would be improved by having higher bandwidth above stock.

What is interesting is that no one seems to have done this test in their reviews yet (or did I miss it?).

What would be even more interesting is if the data proved out this theory as surely AMD would have tested this and determined whether or not bandwidth was really going to be the limitation of the 5870 well before the launch. So why would they artificially handicap the 5870 by knowingly starving it of bandwidth? I don't think we should assume ignorance on their part here.

wouldn't it be better to overclock the memory by intervals as well? say, start at 30% underclock and go up in 5 % increments to as high as you can overclock it.

 

[Wreckage]

Originally posted by: bryanW1995
The real card to compare with gtx 285 is 5850. That card will have all the same features of 5870 but be priced less than gtx 285 by quite a bit.

Quite a bit lower???
http://www.newegg.com/Product/..._-14-143-190-_-Product

 

[dguy6789]

$325 vs $259, seems like quite a bit to me. Even if you want to assume you get the rebate, the 5850 still outperforms the GTX 285 across the board, has full DX11 support, and is cheaper, even if by just a little bit.

 

[Munky]

I'd expect the gap to shrink, maybe even disappear, because Nvidia is increasingly focusing on the GPGPU market, which means a whole bunch of additional transistors which aren't used for gaming.

 

[akugami]

Originally posted by: Idontcare

Originally posted by: MODEL3

Originally posted by: Shaq
Also the FS benchmarks were all at 2560 with 4AA in some cases. It would make more sense to bench at 1680 and 1920 and average the results because the vast majority of gamers have a 22" or 24" monitor. Just showing the highest resolutions and high levels of AA is trying to skew the results in favor of ATI. It is established that the 5870 is better with high AA and high resolutions. But does it really matter if a 285 gets 60 FPS and the 5870 gets 75 FPS in a certain game? They are both very playable. It comes down to cost and availability at that point and for some people power consumption and heat. The 285 can be found for $296 versus the $379 5870 if you can find it. That is 25% cheaper and about how much faster the 5870 is compared to the 285 at common resolutions.

I don't have a problem with 2560 res testing, although i would like 1920 testing also.
For this kind of cards 1680 testing (at launch) is not essential imo.

I don't like 8X AA testing because it favors ATI.
It's like if you want to favor NV, do test at 0X AA.
(of cource i prefer 8X AA testing than 0X AA testing)
The thing is 4X AA produce very good results and the difference between 4X AA and 0X AA is big imo but the difference between 8X AA and 4X AA is much smaller imo regarding quality perception.

---------------------------------------------------------
Regarding prices, you forgot $259 5850.

The obvious answer is you (a responsible reviewer) tests all three combinations at all relevant resolutions so the readers who aim to become buyers have the data that applies to them (resolution and likely AF/AA usage preferences) so they can pare down the data to make just the comparisons they feel represent their likely gaming usage.

People who want to create one-dimensional metrics of comparison (like 3Dmark) are interested in what you write about, and what does practically everyone say about those types of benchmarks?

Originally posted by: Keysplayr
Does quad core benefit in every game? Not really at this point. Maybe many more games next year will be able to utilize as many cores as a system has. That would be excellent.
So, is it safe to say that the 5870 ranges anywhere from 18 to 40% faster than a GTX285 depending on game and res?
Or is this not satisfactory..

:thumbsup:

Having all information on hand so the reader can choose what best fits his or her situation is time consuming but that's IMHO the proper way to do a review. I think there is a point of diminishing returns on testing at lower resolutions such as 1440x900 because pretty much any modern mid to high end video card will perform extremely well at those resolutions.

As for actual testing, especially with different resolutions, we've seen different video card architectures scale up well in higher resolutions and we've seen others that perform well at more mainstream resolutions but don't scale up well at higher resolutions. There are cases where, if you're playing at lower resolutions, Video Card A and Video Card B will work equally well but if you go higher, Video Card A might be better because it scales better and there is less of a hit on performance. Without the full battery of tests and benchmark data we wouldn't know this.

A card like a 5870 can show very very good performance increase compared to a GTX 285 but not in all situations. In some cases it shows substantial performance increases but others, it shows improvements but not as much as we'd all like. To show only the tests with minimal gains shows a bias and an agenda but to show only the tests that shows fantastic gains also shows bias.

Hell, I could make the outrageous claim that the 5870 performs 54% better in GPGPU calculations as based upon this testing of an nVidia ocean demo that utilizes DirectCompute. The number looks gaudy and an actual production app that was optimized for both ATI and nVidia architectures may have each performing within the same neighborhood. If I was trying to be misleading I could certainly make a claim that ATI crushes nVidia in DirectCompute without the proper context and someone who was less informed would be mislead.

 

[SlowSpyder]

Originally posted by: Wreckage

Originally posted by: bryanW1995
The real card to compare with gtx 285 is 5850. That card will have all the same features of 5870 but be priced less than gtx 285 by quite a bit.

Quite a bit lower???
http://www.newegg.com/Product/..._-14-143-190-_-Product

Isn't the 5850 $259? Even if they are the same price, the 5850 out performs the GTX285 and supports DX11. It's really no contest when comparing the two. Nvidia's answer will come with their upcoming GPU's, what they have now isn't meant to compete with the 58x0 cards.

 

[Wreckage]

Originally posted by: SlowSpyder

Isn't the 5850 $259? Even if they are the same price, the 5850 out performs the GTX285 and supports DX11. It's really no contest when comparing the two. Nvidia's answer will come with their upcoming GPU's, what they have now isn't meant to compete with the 58x0 cards.

The 5850 is not available. Whenever it does show up it may be in limited quantities and vary in cost. If you are telling people to wait, then they should wait for a GT300 as well.

DX11? Show me the list of available games. PhysX is far more useful and in far more games right now. DX11 will be more relevant in another GPU generation or 2. Right now it's nothing.

Maybe the GTX2xx series is not meant to compete with the 58xx series, but it does.

The GTX285 performs within about 20% on average (at about $100 less) with a 5870 and a GTX295 beats it.

 

[SlowSpyder]

Originally posted by: Wreckage

Originally posted by: SlowSpyder

Isn't the 5850 $259? Even if they are the same price, the 5850 out performs the GTX285 and supports DX11. It's really no contest when comparing the two. Nvidia's answer will come with their upcoming GPU's, what they have now isn't meant to compete with the 58x0 cards.

The 5850 is not available. Whenever it does show up it may be in limited quantities and vary in cost. If you are telling people to wait, then they should wait for a GT300 as well.

DX11? Show me the list of available games. PhysX is far more useful and in far more games right now. DX11 will be more relevant in another GPU generation or 2. Right now it's nothing.

Maybe the GTX2xx series is not meant to compete with the 58xx series, but it does.

The GTX285 performs within about 20% on average (at about $100 less) with a 5870 and a GTX295 beats it.

I'm not telling anyone to wait, the GTX285 is still a very good card, and unti the 5850 shows up in quanity it can keep it's current price and probably still sell. If someone wants ~GTX285 performance, ~$300 is their budget, and they want to buy right now than the GTX285 is probably a great choice (unless that person doesn't mind multi-GPU, in which case the 4850x2 is a much better choice - cheaper and faster).

There are no DX11 games right now, but just because it's not here right now doesn't mean it's not worth potentially considering... just like the GT300, it's not here now, but for some they may want to consider it and wait on a 5870. Are you suggesting if something isn't here 'right now' it's not worth considering, like the GT300? I would think knowing something is definitely coming and WILL be used it may be worth thinking about in a purhcasing decision, at least with DX11 we know that much. With the GT300 we don't know if it'll be faster, slower, cost, when it'll be out, etc.

The GTX2x0 series competes in some areas, but what would you rather have, a 3870x2 or a 8800GTX? On a strictly performance basis the GTX295 can compete with the 5870, but when you factor in price, power usage, and features I think the 5870 is head and shoulders above Nvidia's offering in this case.

Your 20% number has already been shown to be false, at least at resolutions GTX285 and 5870 level cards are likely to be used.

 

[Idontcare]

Originally posted by: bryanW1995

Originally posted by: Idontcare

Originally posted by: dguy6789
The limitation on the 5870 is most certainly memory bandwidth. Memory bandwidth is the only category in which the 5870 does not outclass the 4870x2 in.

Radeon HD 5870:

Pixel Fill Rate: 27200 MPixels/sec
Texture Fill Rate: 68000 MTexels/sec
Flops: 2720 GFLOPS
Memory Bandwidth: 153.6GB/sec

Radeon HD 4870x2:

Pixel Fill Rate: 24000 MPixels/sec
Texture Fill Rate: 60000 MTexels/sec
Flops: 2400 GFLOPS
Memory Bandwidth: 230.4GB/sec

I would think this theory would be fairly easy to test and either prove or disprove by anyone who has the 5870 in hand.

Just underclock the memory. Do stock, underlocked mem by 10-15%, underclocked mem by 20-30%, and plot the benchmark data.

If you get a linear line with nice R^2 and the slope is nice and large (slope of 1 on a % bandwidth delta versus % fps delta would be perfect) then you can easily conclude the performance would be improved by having higher bandwidth above stock.

What is interesting is that no one seems to have done this test in their reviews yet (or did I miss it?).

What would be even more interesting is if the data proved out this theory as surely AMD would have tested this and determined whether or not bandwidth was really going to be the limitation of the 5870 well before the launch. So why would they artificially handicap the 5870 by knowingly starving it of bandwidth? I don't think we should assume ignorance on their part here.

wouldn't it be better to overclock the memory by intervals as well? say, start at 30% underclock and go up in 5 % increments to as high as you can overclock it.

It would prove the same point, yes.

 

[Tempered81]

yes obv GTX285 > HD5870, whats all the commotion?

 

[dguy6789]

Originally posted by: Wreckage

Originally posted by: SlowSpyder

Isn't the 5850 $259? Even if they are the same price, the 5850 out performs the GTX285 and supports DX11. It's really no contest when comparing the two. Nvidia's answer will come with their upcoming GPU's, what they have now isn't meant to compete with the 58x0 cards.

The 5850 is not available. Whenever it does show up it may be in limited quantities and vary in cost. If you are telling people to wait, then they should wait for a GT300 as well.

DX11? Show me the list of available games. PhysX is far more useful and in far more games right now. DX11 will be more relevant in another GPU generation or 2. Right now it's nothing.

Maybe the GTX2xx series is not meant to compete with the 58xx series, but it does.

The GTX285 performs within about 20% on average (at about $100 less) with a 5870 and a GTX295 beats it.

Are you kidding? The 5870 costs less than 20% more than the GTX 285, but is more than 20% faster.(Your 20% faster average point was debunked already, yet you keep throwing it around)

If you want to go by your logic, the GTX 295 is probably ~20% faster than the 4870 on average if you include resolutions as low as 1024x768, 1280x960, 1440x900, and such in the average. You're the one cherry picking benchmarks. The word "average" does not make your point any more valid.

Where the heck do you get your $100 less nonsense? Newegg shows the cheapest 285 at $325, and the next cheapest at $339. The 5870 is $379. The 5870 has reliably been sold at MSRP, so there is no reason to believe the 5850 will not. There have not been shortage problems either, the 5870 regularly goes in and out of stock on sites as the day goes on.

Comparing PhysX to DirectX 11 is a complete joke, sorry. One is a standard and one is a gimmick. Batman Arkham Asylum is the only game that can be considered not trash that has PhysX in it. We've all been waiting for PhysX to become relevant, it's been years and still a no show. There is no way you can call DX11 useless and PhysX worthwhile in the same sentence with a straight face.

You consistently ignore posts that respond to and disprove your propaganda. You shouldn't be permitted to post if you're going to post incorrect information and not even respond when someone counters it, but continue posting incorrect nonsense.(Trolling)

 

[Wreckage]

Originally posted by: SlowSpyder


I'm not telling anyone to wait, the GTX285 is still a very good card, and unti the 5850 shows up in quanity it can keep it's current price and probably still sell. If someone wants ~GTX285 performance, ~$300 is their budget, and they want to buy right now than the GTX285 is probably a great choice (unless that person doesn't mind multi-GPU, in which case the 4850x2 is a much better choice - cheaper and faster).

On this I can agree.

There are no DX11 games right now, but just because it's not here right now doesn't mean it's not worth potentially considering... just like the GT300, it's not here now, but for some they may want to consider it and wait on a 5870. Are you suggesting if something isn't here 'right now' it's not worth considering, like the GT300? I would think knowing something is definitely coming and WILL be used it may be worth thinking about in a purhcasing decision, at least with DX11 we know that much. With the GT300 we don't know if it'll be faster, slower, cost, when it'll be out, etc.

I kind of disagree on this. DX10 is just now starting to gain ground as DX11 cards show up. I'm betting it will be a while (3+ years) before DX11 is of any real need. Physx as I noted is far more useful and in far more games right now.

The GTX2x0 series competes in some areas, but what would you rather have, a 3870x2 or a 8800GTX? On a strictly performance basis the GTX295 can compete with the 5870, but when you factor in price, power usage, and features I think the 5870 is head and shoulders above Nvidia's offering in this case.

I'm not a big fan of multi GPU so I will not argue your statement here. My only point was that the GTX2xx series does actually compete with the 5870 series. 1 on price the other on performance.

Your 20% number has already been shown to be false, at least at resolutions GTX285 and 5870 level cards are likely to be used.

It has been mostly attacked by cherry picking benchmarks, so I would not say it has been proven false.

 

[Wreckage]

Originally posted by: dguy6789
Are you kidding? The 5870 costs less than 20% more than the GTX 285, but is more than 20% faster.(Your 20% faster average point was debunked already, yet you keep throwing it around) Where the heck do you get your $100 less nonsense? Newegg shows the cheapest 285 at $325, and the next cheapest at $339. The 5870 is $379. The 5870 has reliably been sold at MSRP, so there is no reason to believe the 5850 will not. There have not been shortage problems either, the 5870 regularly goes in and out of stock on sites as the day goes on.

:roll: Of course you ignore the rebate. :roll:

Here's another one.
http://www.microcenter.com/sin...tml?product_id=0317763

Comparing PhysX to DirectX 11 is a complete joke, nice try. Batman Arkham Asylum is the only game that can be considered not trash that has PhysX in it. We've all been waiting for PhysX to become relevant, it's been years and still a no show. There is no way you can call DX11 useless and PhysX worthwhile in the same sentence with a straight face.

At least show me a list of how many games are on the market using Physx and how many are using DX11.


No reason to get angry and offensive over this...sheesh.

 

[dguy6789]

The rebate is a gamble, it can not be relied on. Even if it was guaranteed, it's still not a $100 difference. Three DX11 games are supposed to come out Q4 this year.

 

[Wreckage]

Originally posted by: dguy6789
The rebate is a gamble, it can not be relied on. Even if it was guaranteed, it's still not a $100 difference.

The microcenter card is $279 AR. How much are 5870s?

Three DX11 games are supposed to come out Q4 this year.

In other words none.

 

[dguy6789]

Ah I didn't see the rebate on the Microcenter link. Fine then, the GTX 285 can be had for $100 less than the 5870. That does make it a reasonable purchase.

 

[Wreckage]

Originally posted by: dguy6789
Ah I didn't see the rebate on the Microcenter link. Fine then, the GTX 285 can be had for $100 less than the 5870. That does make it a reasonable purchase.

See. Can't we all just get along?

The 5870 is a great card. AMD did a good job with it. I was just discussing certain things I have read that helped me to form an opinion regarding the 5870 vs the GT300. (what this thread is about).

Until the GT300 shows up it's all just make believe anyways.