Are next gen highend GPU's going be priced like the Titan ??

[Rusty0611]

You'll have 6GB now you greedy bugger anyway .

did we get abit off topic here?

Im all for a great debate but i got lost after the first wall of text

Yes just a little . It's just Tommy has been chuntering about this in multiple threads for a few days (and I'm not the only one who has noticed) and I was fed up of reading it.

 

[LtMatt]

Me and PGI tested Lightnings. Remember it was a clock for clock test which a few including me wanted to see. Hopefully Paul will show the chart but not fussed either way.

As for VRAM, it wouldn't matter if I had the 4GB cards because they simply can't cope with decent/acceptable frame rates at 5760*1080 with all the bells and whistles on.

Whatever way you look at it now, the 7970Ghz edition is faster than the 680. Clock for clock, the 7970 is faster than the 680. 2*7950's are faster than 2*680's at 5760*1080.

Anyways, let's get back on topic and have a group hug

Can't argue with that.

Though that comment in the msi afterburner thread has ruffled my feathers. However seeing as you didn't mention the holy grail of amd cards (you know ) im willing to let it slide this once.

 

[Gregster]

Can't argue with that.

Though that comment in the msi afterburner thread has ruffled my feathers. However seeing as you didn't mention the holy grail of amd cards (you know ) im willing to let it slide this once.

 

[tommybhoy]

It was close in April/May but it wasn't official until November?

Now I'm truly lmfao, what was the point of arguing that performance never arrived until November then?

All I said was performance has always been available, now I'm getting hit with 'incorrect absolutist statements' when there were none from me.

Turn it around and apply that to your 12.11 drivers was when AMD got a grip statements.

No one here is telling me I'm wrong here bar you rusty.

 

[Rusty0611]

Though that comment in the msi afterburner thread has ruffled my feathers.

Good banter. I'll leave you on my bookmarks for now.

I had to dig your guide out when I got my second 7950 as I forgot how to unlock the clock limits so I stuck it on there for future reference. It took so long, after having to sift through your 2000+ gif posts that I almost threw the towel in....

 

[Rusty0611]

It was close in April/May but it wasn't official until November?

Now I'm truly lmfao, what was the point of arguing that performance never arrived until November then?

Give it up, it's becoming painful now. That just doesn't make any sense. You're going through my post, reading it, turning it around in your head to try and find something disagreeable.

All I said was performance has always been available

Well that just doesn't mean anything....

No one here is telling me I'm wrong here bar you rusty.

Well actually Gregster is as well but OK... not as bluntly as me but the end point is the same.

Also, I'm not sure anybody else is even reading this. It's a bit boring and I know if I wasn't involved in the discussion I would have skipped over it.

 

[LtMatt]

What are you/greg saying he's wrong over Rusty?

 

[Rusty0611]

VRAM / performance pre/post 12.11 while overclocked.

I'm not that interested in stock performance as there's a plethora of reviews already on the net showing these over time.

Edit: we know that the GHz edition launch re-took the "fastest GPU crown". I don't think anyone is really debating this... or at least I hope Tommy hasn't got his knickers in a twist and is debating this.

 

[tommybhoy]

Greg's only talking about 1 point out of three it's progressed to.

There is agreement with the other two original points rusty or do you not compute other opinions bar what you see fit?

@Matt, I was talking about max oc performance from launch, GHz only comes into it when the majority of reviewers pointed out the GHz was =/touch faster@launch, in June.

 

[Rusty0611]

Greg's only talking about 1 point out of three it's progressed to.

Tommy on VRAM - wrong.

As for VRAM, it wouldn't matter if I had the 4GB cards because they simply can't cope with decent/acceptable frame rates at 5760*1080 with all the bells and whistles on.

Tommy on OC performance prior to 12.11 - wrong.

Me and PGI ran BF3 clock for clock prior to 12.11 and then again after. In this game, BF3 was faster on the 680 till the 12.11. This was at the most used resolution of 1080P.

The 12.11 drivers was the only time the 7970 pulled ahead of the 680.

Tommy on 10% - inconclusive.

While your evidence would be useful in addition to other information, on its own it just doesn't mean anything.

 

[LtMatt]

VRAM / performance pre/post 12.11 while overclocked.

I'm not that interested in stock performance as there's a plethora of reviews already on the net showing these over time.

Edit: we know that the GHz edition launch re-took the "fastest GPU crown". I don't think anyone is really debating this... or at least I hope Tommy hasn't got his knickers in a twist and is debating this.

I remember PGI+Greg running a couple of benchmarks on a couple of games where by they used the same clocks. I remember bf3 was one of them and gregs was slightly faster but that was during the time where nvidia had the upper hand with bf3. The problem was though in that game the testing method was in no way reliable. I'd certainly be interested in running some tests like this again in a few titles if we have any 680 users willing to partake. We need games that have a solid and reliable benchmarking system though like sleeping dogs. Not one where you play a part of a single player mission as its hard to draw comparisons from that when different things can happen.

EDIT

Would be handy if someone still had that benchmark greg+pgi did.

 

[tommybhoy]

Tommy on VRAM - wrong.

Tommy on OC performance prior to 12.11 - wrong.

Tommy on 10% - inconclusive.

While your evidence would be useful in addition to other information, on its own it just doesn't mean anything.

1st one, jury's out, but reports are coming that the 2GB is choking@1080p in shader intensive titles-vram/memory bus bottleneck

2nd one, evidence has been about since April that shows it was more or less equal

3rd one lmfao, it's only wrong because you didn't test it mate, it would be law if you did.

 

[Rusty0611]

I remember PGI+Greg running a couple of benchmarks on a couple of games where by they used the same clocks. I remember bf3 was one of them and gregs was slightly faster but that was during the time where nvidia had the upper hand with bf3. The problem was though in that game the testing method was in no way reliable. I'd certainly be interested in running some tests like this again in a few titles if we have any 680 users willing to partake. We need games that have a solid and reliable benchmarking system though like sleeping dogs etc.

It was the same benchmark review sites used so it was only as reliable as the baseline set by them. I agree there will be differences but due to the benchmark being quite long it is unlikely to affect the average FPS by more than 0.25 either way.

 

[LtMatt]

It was the same benchmark review sites used so it was only as reliable as the baseline set by them. I agree there will be differences but due to the benchmark being quite long it is unlikely to affect the average FPS by more than 0.25 either way.

Its not reliable enough in my opinion. I posted a fair few benchmarks from that section in the ocuk benchmark thread and my scores did change a fair bit. Sometimes more than one car would blow up which could hurt the fps. Sometimes you'd be zoomed in using the rifle and would miss the explosion giving your fps a nice gain. Too many variables even in a brief scene like that. Consistency is key in these things in my opinion.

EDIT

From memory there's a better benchmarking part in the single player campaign called "Going Hunting" which is far more consistent. The fps in this scene are much higher though.

 

[Rusty0611]

1st one, jury's out, but reports are coming that the 2GB is choking@1080p in shader intensive titles-vram/memory bus bottleneck

So you've gone from using Gregster as an example, except he's said you're wrong, to using "reports". i.e. not a lot of substance

2nd one, evidence has been about since April that shows it was more or less equal

No, the benchmark thread says otherwise and user results are for more believable that review sites IMO. Do you really want me to dig out a post from you where you said that exact thing? But not because it suits you've changed your mind?

3rd one lmfao, it's only wrong because you didn't test it mate, it would be law if you did.

You can't draw conclusions like you have based on a synthetic benchmark which may/may not have any relevance to games. I'd say you were being disingenuos but you probably don't actually see the problem with it.

 

[Rusty0611]

Its not not reliable enough in my opinion. I posted a fair few benchmarks from that section in the ocuk benchmark thread and my scores did change a fair bit. Sometimes more than one car would blow up which could hurt the fps. Sometimes you'd be zoomed in using the rifle and would miss the explosion giving your fps a nice gain. Too many variables even in a brief scene like that. Consistency is key in these things in my opinion.

I agree you can't get identical runs but it may change your minimums but averages won't change much at all. If it does then something else is causing the changes just purely down to the length of the benchmark.

Unfortunately, it's the only way to really bench BF3 as close to consistently as you can and getting the key moments in. This is why review sites use this.

 

[LtMatt]

1st one, jury's out, but reports are coming that the 2GB is choking@1080p in shader intensive titles-vram/memory bus bottleneck

2nd one, evidence has been about since April that shows it was more or less equal

3rd one lmfao, it's only wrong because you didn't test it mate, it would be law if you did.

Not sure about the first point. The second point i can agree with to some degree. The catchup had started before Ghz edition came out. 3rd point i agree with. I don't think you can apply a one size percentage fits all to this. Its between 5-10% depending on the title.

I seriously think we should settle this via benchmark in the ocuk benchmark thread. I for one think it would be interesting and love a good bar chart. I can't remember that last time we posted some benchmarks either.

 

[Boomstick777]

Agreed. I expect the 780 from rumours to be around 10% slower than a Titan at best. Maybe the 8970 will be a little closer, as again rumours were touting 25% faster. I am only going on rumours though and this could be miles off. Maybe AMD want the crown back, so anything is possible.

I think the 8970 may be a lil faster than the Titan at gaming, but the Titan will still retain value in resale etc due to it's parallel computing and GPGPU technology.. Useful for researchers and other professionals that need compute performance.

We're paying the extra for having 8970 / 780 performance now rather than at the end of the year, so even if AMD launch a 8970 that's faster doing it 6 months later @ $599 would be expected.. I'm using 1440P screen so Titan will be ideal right now..

http://images.anandtech.com/graphs/graph6773/53222.png

** No Hotlinking - Rules on posting images ***

 

[tommybhoy]

So you've gone from using Gregster as an example, except he's said you're wrong, to using "reports". i.e. not a lot of substance

My reference to greg and the quote was in the same said post.

I'm wrong in a sense that i can't prove that 7970 CrossFire can play titles better than a 2GB card on a smaller memory bus, then again my proof counts as nothing.

No, the benchmark thread says otherwise and user results are for more believable that review sites IMO. Do you really want me to dig out a post from you where you said that exact thing? But not because it suits you've changed your mind?

Dig out posts if you like, I'm in agreement that user data is the best bet but there was no BM thread in April/May so all there is to go on is review data-which has been posted and obviously ignored by some.

What data did you use to purchase your 680?

You based your purchase on reviews, didn't you?

You can't draw conclusions like you have based on a synthetic benchmark which may/may not have any relevance to games. I'd say you were being disingenuos but you probably don't actually see the problem with it.

I/others don't see a problem with the methology, only yourself.

@Matt, there is no chance my lazy 7950 is going in the BM thread, rusty would fall about laughing if I did due to it's poor results.

 

[LtMatt]

You're going to be our 7950 user and I'm going to be the 7970 user Tommy. No getting out of it. I'm already making plans and drawing up some games/benchmarks we can run as a test. I'm thinking one set at 7950 & 7970 stock clocks. Then another run where by we both overclock and clock match. Depending on how well your measly 7950 can keep up with my elitist msi 7970 oc we could run 1125/1575.

Then we just need two Nvidia users to do the same for the other side with a 670+680.