AMD: R9xx Speculation

[NathansFortune]

Hey, I'm just quoting the guy who has one. It's his benchmark result. Also he has a reference 6970 which has no OC, if you have an OC for your 5870 it could affect your results.

 

[neliz]

And what about the famous "switch"?

The switch is a setting for bios upgrades, it's been explained a trazillion times already.

10.11 borks performance (even on the 6850/70)

You need something better than the latest publicly released 10.11 beta's.

 

[CarstenS]

Take a close look at the 3DMark 11 results. The driver version's given there.

 

[leoneazzurro]

The switch is a setting for bios upgrades, it's been explained a trazillion times already.

10.11 borks performance (even on the 6850/70)

You need something better than the latest publicly released 10.11 beta's.

Sorry, it was only said it was a "Dual BIOS " switch, but for not what that meaned.

 

[NathansFortune]

Take a close look at the 3DMark 11 results. The driver version's given there.

8.790.6.0

Is that out dated?

 

[ECH]

I assumed that 8.790.60 is what came on disc. The official 10.11 WHQL driver is 8.791.0.0. While Cat 10.10e uses 8.782.1. New cat 10.11 uses version 8.79.6.1.0

 

[Alexko]

It is, when you click on go to full article below. Word for word the same.
So, I wonder, what did keep Charlie from taking a picture while making his caliper measurements. It really isn't that hard.

I don't think he made the measurements himself. Actually, he probably doesn't have the card either.

 

[ECH]

F1 2010 results

GTX580 @ 1920x1080
8x AA
65 avg
76 max
59 min
http://www.hardwareheaven.com/revie...vidia-geforce-gtx-580-sli-review-f1-2010.html


hd6970 @ 1920x1080
8x AA
68 avg
58 min
Samples: 8554

 

[Picao84]

F1 2010 results

GTX580 @ 1920x1080
8x AA
65 avg
76 max
59 min
http://www.hardwareheaven.com/revie...vidia-geforce-gtx-580-sli-review-f1-2010.html


hd6970 @ 1920x1080
8x AA
68 avg
58 min
Samples: 8554

This is no better result than an HD5870 with 2 GB RAM, according to the HardOCP review...

 

[Xenus]

Unfortunately with the different bus sizes it's going to be hard to tell when a card is ram limited though I guess if they eventually come out with a 580 3GB we might be able to tell there.

 

[ECH]

This is no better result than an HD5870 with 2 GB RAM, according to the HardOCP review...

Hmm, I don't recall them using a 5870 @ 1080 resolution for any F1 2010 benchmark

 

[Picao84]

Hmm, I don't recall them using a 5870 @ 1080 resolution for any F1 2010 benchmark

No, but they used 2560x1200 8xAA, 16AF

Link: http://www.hardocp.com/article/2010/11/09/nvidia_geforce_gtx_580_video_card_review/3

The aging ATI Radeon HD 5870 seems to hold its own in this game, and compares right up there with the new GeForce GTX 580. We found that 8X MSAA at 2560x1600 was playable on both the GeForce GTX 580 and the ATI Radeon HD 5870.

 

[mao5]

This is no better result than an HD5870 with 2 GB RAM, according to the HardOCP review...

6970 42fps 5870 2GB 38fps, 580 39.4 fps, 580 is dead in F1 2010 2560x1600 8xMSAA

let alone the German guy's testing platform is not very well

 

[Picao84]

6970 42fps 5870 2GB 38fps, 580 39.4 fps, 580 is dead in F1 2010 2560x1600 8xMSAA

WOW 3 -4 FPS faster!!! I bet that can change your gaming experience... NOT!
This only might prove that both cards are hitting memory limitations, which does not say anything to us about the true power of these solutions. So it might not be a good term for comparison.

(Note: Not saying this like an nVIDIA fanboy, but if Cayman is not much more quicker than a 2GB RAM HD5870, there must be a bottleneck, which might be the memory, IMO)

 

[ZerazaX]

Hards method doesnt help for apples to apples comparison - and the resolution is different too, the 2GB on the 5870 might be saving it at 2560x1600

1920x1080 is where an apples to apples comparison can be made w/ those scorse

 

[Picao84]

Hards method doesnt help for apples to apples comparison - and the resolution is different too, the 2GB on the 5870 might be saving it at 2560x1600

1920x1080 is where an apples to apples comparison can be made w/ those scorse

They do have a graph with apples to apples

Apples-to-Apples

In this apples-to-apples test we are running each video card at 8X MSAA along with their respective multisample AA methods.

Care to look at the LINK, before stating that, please?

 

[ZerazaX]

WOW 3 -4 FPS faster!!! I bet that can change your gaming experience... NOT!
This only might prove that both cards are hitting memory limitations, which does not say anything to us about the true power of these solutions. So it might not be a good term for comparison.

(Note: Not saying this like an nVIDIA fanboy, but if Cayman is not much more quicker than a 2GB RAM HD5870, there must be a bottleneck, which might be the memory, IMO)

No, you're using the wrong methodology. You can't compare the cards using [H]'s results because:

[H] looks at highest settings - you have to look at apples vs apples in THE SAME resolution to compare results - for all we know, the 580 trounces the 5870 2GB in 1920x1080 but runs into frame buffer limitations at 2560x1600

So unless the 5870 2GB is beating it in 1920x1080, which [H] doesnt show, your argument is invalid since THE RESOLUTION is different!

 

[Picao84]

No, you're using the wrong methodology. You can't compare the cards using [H]'s results because:

[H] looks at highest settings - you have to look at apples vs apples in THE SAME resolution to compare results - for all we know, the 580 trounces the 5870 2GB in 1920x1080 but runs into frame buffer limitations at 2560x1600

For ****** sake people, do you care at looking at the link or not?
They do have an Apples to Apples comparison!

 

[ZerazaX]

For ****** sake people, do you care at looking at the link or not?
They do have an Apples to Apples comparison!

STOP ignoring what I said? Where is the Apples to Apples for 1920x1080?

You're seriously missing out on scientific method 101 here by including a seriously big variable

2560x1600 is 4MP, 1920x1080 is 2MP, BIG difference

 

[Picao84]

STOP ignoring what I said? Where is the Apples to Apples for 1920x1080?

You're seriously missing out on scientific method 101 here by including a seriously big variable

2560x1600 is 4MP, 1920x1080 is 2MP, BIG difference

Still, you can see that this particular game IS very memory intensive, because on other games, with similar settings, the GTX580 trounces the HD5870. Then, I still think its NOT a good term of comparison, which was my stance after all.