Poll: GT300 VS. HD5870

[akugami]

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.

Double standards. Many didn't mind when nVidia released video cards in the past to limited availability with etailers constantly sold out and the actual cost about $30 higher than the MSRP.

Games coming soon with DX11 support I don't know the exact number of PhysX hardware accelerated games but DX11 will equal, or surpass, the number of PhysX hardware accelerated games in it's first year of availability. DX11 is coming. It's not here now, it's not even relevant yet and likely won't be relevant until 2011 but it's coming.

DX11 is more of a sure thing than PhysX. In fact, DX11 success is pretty much guaranteed since even if a DX12 is released, it'll likely be a superset of DX11 that expands on the direction DX is going. Assuming games don't get delayed there will be Alien VS Predator, Dirt2, Battleforge, Dungeons and Dragons Online, LOTR Online, STALKER: Call of Pripyat by Q1 of next year. Which if you're looking for a card that supports DX11 means you should wait for the GT300 and make a comparison on what to buy then.

There has been more than one comment on the 20% numbers you claim between the GTX285 and Radeon 5870. It just shows your bias more than ever. Even after people show how you skewed your data to represent what you want. Because everyone knows the Radeon 5870 is twice as powerful as a GTX 295 in GPGPU calculations as shown in early DirectCompute testing. On an nVidia released app no less. It's also close to 40% more powerful than a GTX 285 in many games. I can also say that the 5870 can be found for $379.99 while the GTX 285 sells for $397.99 on Newegg. :roll:

Was "the facts" I stated above the truth? Yes, but it's also skewing the data to fit my point. If I was trying to show ATI in as good a light possible I'd make those claims without any relevant context. Just like you are showing your bias against ATI with the data you represent. As the saying goes, "There are three kinds of lies: lies, damned lies, and statistics."

Bottom line, the Radeon 5870 is a solid upgrade vs the 4870 and GTX 285. Would we love more performance out of it? Sure, but the same can be said of any new ATI or nVidia video card that is released.

 

[RussianSensation]

Originally posted by: Wreckage

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

In other words none.

Battleforge and Dirt 2 are both out and are both DX11 capable.

5870 is 80% faster than a 285 in Battleforge when DX11 is enabled.
1920x1200 4AA
5870 = 38 avg / 28 min (+81% avg / +87% min)
4890 = 25 avg / 18 min
285 = 21 avg / 15 min (Base = 100)

 

[Qbah]

Originally posted by: Wreckage

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.

What cherry picking? I clearly showed that this chart if full of bull. First you can't properly read it (the GTX285 is 18.4% slower and not the other way round, there's a difference) - the HD5870 is 22.5% faster than a GTX285 from looking on the chart.

Second, the same chart shows that a GTX275 SLI is only 20% faster than single GTX275 :disgust: And the GTX275 SLI setup is slower than a single HD5870 (by 8%). Yet somehow a GTX285 in SLI is 16% faster.

You're basing your claim on a chart that has glaring errors in the first place. Then you can't properly read it. It also doesn't explain how this average was created. It links to articles, that don't have meaningful data or no data at all. So how can you keep repeating your false 18% even though it was proved it's wrong? That's right - you're trolling.

 

[GaiaHunter]

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.

Actually it doesn't, since the 4890 according to the presented charts that leads to the "20% advantage of the 5870 over the GTX 285" is only 7% slower than the GTX 285 and can be had for $189.

 

[Wreckage]

Originally posted by: akugami


Games coming soon with DX11 support I don't know the exact number of PhysX hardware accelerated games but DX11 will equal, or surpass, the number of PhysX hardware accelerated games in it's first year of availability. DX11 is coming. It's not here now, it's not even relevant yet and likely won't be relevant until 2011 but it's coming.

Many of those say TBA. Either way still none.

Here's a list of Physx titles.
http://www.nzone.com/object/nzone_physxgames_all.html

DX11 is more of a sure thing than PhysX.

Eventually maybe, but not for years.

There has been more than one comment on the 20% numbers you claim between the GTX285 and Radeon 5870. It just shows your bias more than ever.

My bias? I did not write the article or generate the graph. Nor did I run any of the benchmarks the site used to get to that number.

Bottom line, the Radeon 5870 is a solid upgrade vs the 4870 and GTX 285.

I disagree 100%.

Originally posted by: RussianSensation

Battleforge and Dirt 2 are both out and are both DX11 capable.

Could you link to where I could buy dirt 2?

 

[uclaLabrat]

Originally posted by: Wreckage

Could you link to where I could buy dirt 2?

Here
you go

 

[imported_Shaq]

That is a coupon. He means to buy and play it today. I don't even see a listing for DIRT2 on Ebgames. If it is included it makes Batman and PhysX exclusivity pale in comparison.

 

[uclaLabrat]

Originally posted by: Shaq
That is a coupon. He means to buy and play it today.

I was kidding. Thanks for playing!

 

[Schmide]

People please be careful of x% faster and x% slower. If you talk about speeds be consistent, all faster or all slower.

Example: card X is 50% faster than card Y , means Card Y is 33.3% slower than card X.

 

[Qbah]

Originally posted by: Schmide
People please be careful of x% faster and x% slower. If you talk about speeds be consistent, all faster or all slower.

Example: card X is 50% faster than card Y , means Card Y is 33.3% slower than card X.

It seems "some" of the posters here can't grasp this simple logical difference.

 

[GaiaHunter]

Originally posted by: Shaq
If it is included it makes Batman and PhysX exclusivity pale in comparison.

Batman is another example of cutting simple physics effects other games already did using cpu resources to make the physX present a much higher visual difference than otherwise.

PhysX is quite a poor argument in my opinion...

Not that I would buy a 5870 if I had a 4870/GTX 260 or better now.

But I wouldn't buy a GTX 285 or GTX 295 now either.

 

[GaiaHunter]

Originally posted by: Qbah

Originally posted by: Schmide
People please be careful of x% faster and x% slower. If you talk about speeds be consistent, all faster or all slower.

Example: card X is 50% faster than card Y , means Card Y is 33.3% slower than card X.

It seems "some" of the posters here can't grasp this simple logical difference.

Or it can just be really annoying to have to keep doing conversions. Maybe we should talk in % of performances in comparison to a set card, in this case the 5870.

Like this having 81% of the 5870 means that the GTX 285 is 19% slower or that the 5870 is 25% faster than the GTX 285.

 

[Schmide]

Originally posted by: Qbah

Originally posted by: Schmide
People please be careful of x% faster and x% slower. If you talk about speeds be consistent, all faster or all slower.

Example: card X is 50% faster than card Y , means Card Y is 33.3% slower than card X.

It seems "some" of the posters here can't grasp this simple logical difference.

Maybe other posters use, cough cough, this technique to seemingly bias their argument.

 

[rpglord]

I cant belive how some people just make generalisations which are not true.
For example,several people on this thread have mention that if you game at 1680 res you dont really need anything faster then 275/280/4890.
Well thats complete bs. I game at that res,and there is dozen of games which works slower then 30 fps ( not to mention 60 fps,to archive that not even new radeon 5870 is enough )
I have c2q at 3,6 ghz and 285 gtx...so I dont need anything more since I game at 1680 right ? Everything more is overkill ?
Crysis at max settings with 4x AA ( I dont dare go higher ) : avarage 28 fps minimum 21
Armad Assault 2 : max setting except view distance : 24 fps avarge minimum 13 fps
Armed Assault 2 : max settings : 10 fps
Far Cry 2 with max setting and 4x AA : around 35 fps
NWN 2 - 33 fps
Etc etc....as you can see these games are not even close to 60 fps,and ARMA2 and Crysis even go below 30 fps ( arma 2 with all settings max is 10 fps )
So somebody can explain me again how is it that I dont need anything faster then 285 ???
I am pretty sure not even radeon 5870 is fast enough at 1680 res....

 

[Idontcare]

Originally posted by: Schmide

Originally posted by: Qbah

Originally posted by: Schmide
People please be careful of x% faster and x% slower. If you talk about speeds be consistent, all faster or all slower.

Example: card X is 50% faster than card Y , means Card Y is 33.3% slower than card X.

It seems "some" of the posters here can't grasp this simple logical difference.

Maybe other posters use, cough cough, this technique to seemingly bias their argument.

Hanlon's razor

Never attribute to malice that which can be adequately explained by stupidity.

 

[Golgatha]

Originally posted by: rpglord
I cant belive how some people just make generalisations which are not true.
For example,several people on this thread have mention that if you game at 1680 res you dont really need anything faster then 275/280/4890.
Well thats complete bs. I game at that res,and there is dozen of games which works slower then 30 fps ( not to mention 60 fps,to archive that not even new radeon 5870 is enough )
I have c2q at 3,6 ghz and 285 gtx...so I dont need anything more since I game at 1680 right ? Everything more is overkill ?
Crysis at max settings with 4x AA ( I dont dare go higher ) : avarage 28 fps minimum 21
Armad Assault 2 : max setting except view distance : 24 fps avarge minimum 13 fps
Armed Assault 2 : max settings : 10 fps
Far Cry 2 with max setting and 4x AA : around 35 fps
NWN 2 - 33 fps
Etc etc....as you can see these games are not even close to 60 fps,and ARMA2 and Crysis even go below 30 fps ( arma 2 with all settings max is 10 fps )
So somebody can explain me again how is it that I dont need anything faster then 285 ???
I am pretty sure not even radeon 5870 is fast enough at 1680 res....

Yes, I too love the folks who think a 22in monitor at 1680x1050 resolution can't utilize two graphics cards or one 5870. I've ran dual graphics cards for quite awhile now and there are plenty of games out there that can slow down even a dual graphics setup with the right combination of in-game and driver level quality settings.

 

[Tempered81]

Originally posted by: Wreckage

There has been more than one comment on the 20% numbers you claim between the GTX285 and Radeon 5870. It just shows your bias more than ever.

My bias? I did not write the article or generate the graph. Nor did I run any of the benchmarks the site used to get to that number.

Is that the chart where the 8800gt is 20% faster than the 9800gt? If so, would make sense that 5870 is only 20% faster than a gtx285.
:roll:

 

[alcoholbob]

Even though it's not advertised, the XFX 5870 also comes with Dirt 2. My guess is other OEMs probably also bundle it; it might be standard to the package so nobody bothers to report it.

 

[Kuzi]

I checked the xbitlabs review here, and added up the performance difference between the HD5870 vs the 4870 x2 and GTx295 cards. For H.A.W.X I used anandtech?s numbers because xbitlabs used Adaptive AA which is incompatible with that game and lowers the framerate on ATI cards. Below is how the HD5870 compares in the 13 games tested by xbitlabs:

Overall HD5870 @ 1680x1050 AA4x / AF16x:
+2% faster than HD4870 x2
-0.5% slower than GTX 295

Overall HD5870 @ 1920x1200 AA4x / AF16x:
+2.2% faster than HD4870 x2
+0.7% faster than GTX 295

Overall HD5870 @ 2560x1600 AA4x / AF16x:
+1.4% faster than HD4870 x2
-2.7% slower than GTX 295

Seems like the difference between the three cards is very minimal at all resolutions, and really depends on which games you're playing. To me the HD5870 is a great card especially when taking into account it's higher OC ability, DirectX 11 support, no microstutter, and lower prices. And with newer future drivers, I see the HD5870 gaining another 5-10% performance on average, which means it will be ahead of those older dual-GPU cards.

 

[Kuzi]

As for the GT300 vs HD5870, I think the GT300 will be around 20-30% faster, and I believe it would still beat a future HD5890 @1GHz GPU/1400MHz GDDR5 by a bit, but will end up slower than an HD5870 x2.

 

[alcoholbob]

The wikipedia specs say it's got theoretically 37% more power than a 295. Now performance doesn't really scale linearly with more horsepower but 30% faster than HD5870 (with 4xAA) sounds likely.

 

[Rubycon]

Whatever they do I wish they would improve the desktop performance of these cards for folks running 2560x1600x32 and higher.

 

[x3sphere]

I have a feeling the GT300 will surprise us. To what extent who knows but I'm thinking along the lines of the 7XXX to 8800GTX jump, so around 30-40 percent performance increase over the 5870. That's why I'm not jumping on a 5870 just yet. By the way, the GT300 is said to be based on a completely new architecture so this kind of performance jump is definitely within the realm of possibility.

 

[swing848]

Idontcare,

Every time ATI releases a video card the drivers take time to mature, this is no exception and I doublt that there is an conspiracy.

ATI is usually behind Nvidia in marketing ability, this time they need to step it up and get software developers lined up, and from what I have read they are already writing code for the HD5870. One reason is the product is on the market, the other reason is the HD5870 is already as fast or beating Nvidea's 295 dual GPU in many games. Unfortunately for Nvidea's next generation product is still vaporware, with no specific release date; December 09 to 1Q 2010 is not very specific.

Developers will write code for hardware that is here, especially if it is fast.

 

[swing848]

Tempered81,

Comparing hardware that is on the market [HD5870] to vaporware from Nvidia is speculation at its best. I have no idea who the winner will be. It is not in Nvidia's favor that software writers are coding for ATI because they have a good product on the market that will get better over time with better drivers and X2 versions of their cards.

Nvidia has good marketing and they will need it in this case. Nvidia still has to overcome major manufacturing problems and when they eventually have a viable product good drivers will have to be written.

I hope Nvidia does a good job with their GT300, however, given the above, at this date, I do not know how anyone can make any kind of prediction regarding performance.

Edit:

ATI was smart in working out the bugs in their HD4770 long before the release of the HD58XX series, they took a page from Intel's play book of putting next generation CPUs in laptops, and working out the bugs before releasing hardware to consumers using desktops and servers.