Besides your obvious points being "obvious", You do have to take a cross the board average to compare cards...I'm still having a problem with 95% better in anything !, you saying BF3 ?
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Geforce Titan rumours.
- Thread starter Gregster
- Start date
Besides your obvious points being "obvious", You do have to take a cross the board average to compare cards...I'm still having a problem with 95% better in anything !, you saying BF3 ?
Soldato
, Think You could have worked out I called SLI, "Crossfire", slight freudian slip there...my error
That's not a Freudian slip.
A Freudian slip would be, for example, when you say to your wife, "why don't you **** off you *****, you've ruined my life", when you meant to say, "can you pass the salt please?".
Haha, funny.
Of course I knew what you meant but doesn't mean you weren't wrong.
It seems BF3 does scale like you say , but we are gonna have to beg to differ over a direct card to card % comparison
My stance on 2 card setups was an overwhelming disappointment and unless I am forced into requiring the extra grunt for a special setup in the future, I intend to avoid
That is why I have an interest in this new card.
Soldato
It seems BF3 does scale like you say , but we are gonna have to beg to differ over a direct card to card % comparison
My stance on 2 card setups was an overwhelming disappointment and unless I am forced into requiring the extra grunt for a special setup in the future, I intend to avoid
No lol. You can't take an across the board average for the reasons which I have stated above. It distorts the picture completely.
Fine but as I have said all along 45% faster than a 680 is grossly incorrect. The only reason it would be so is due to the software being used to test this but this is a fault of the software as opposed to actually being only 45% faster in real terms.
Soldato
More rumour here!
Soldato
Link is dog.
They're talking about the memory and comparing it to 690/7990 4GB/6GB and pointing out that the Titan has 6GB as well but marking it as a difference by saying the 690/7990 are "non-reference" when really the effective VRAM is 2GB/3GB anyway as they're both dual GPU cards and Titan is single.
No lol. You can't take an across the board average for the reasons which I have stated above. It distorts the picture completely.
Fine but as I have said all along 45% faster than a 680 is grossly incorrect. The only reason it would be so is due to the software being used to test this but this is a fault of the software as opposed to actually being only 45% faster in real terms.
You in Advertising ?All the time I used to run 2 card setups I used to say to myself, What's the problem here, first card renders a frame then the second card renders a frame,
why does it not always give me a big boost ! can not be to hard for those driver wizz kids to sort out....but it never happened, when it did the benchmarks looked good but the games never played to expectations.
Now I have not used 2 cards for a couple of years, but what I see around forums is the same issues, sure maybe when they work well, they are better than than they used to be, but they are not there yet !
The point is You just can not say a card is 95% better "PROVIDING" You only run these games !
This new Titan we should be able to says is X% better across the full board
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Firstly, it's not crossfire.
Secondly, in benchmarks I got near perfect scaling going from one GPU to two and in games like BF3 the same was observed.
Thirdly, it just is around 95% faster - 90% at worst - so if you're surprised this says more about you than it does my percentages.
It's funny that you've gone around saying its disingenuous to say the 690gtx is only 45% faster, while also stating that the 690gtx is 90% faster at worst, which is entirely bullcrap on its own.
http://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_690/14.html
It's not even 50% faster at 1920x1200, and it stretches that to around 75% faster at 2560x1600 or surround res. Neither of these is 90%, or close to it.
http://www.anandtech.com/show/5805/...eview-ultra-expensive-ultra-rare-ultra-fast/5
The 690 is close to but under 80% in most resolutions there, its under 80% in BF3 at all resolutions except 1920x1080 with MSAA, where its just above 81%. Metro is around 55-65% faster, Dirt 3 is around 80%, Shogun actually has exceptional scaling, though its around 100% for AMD as well, and the 7970's demolished Nvidia with ultra quality on, that will almost certainly be down to bandwidth and one of the places Titan should do the best.
http://www.guru3d.com/articles_pages/geforce_gtx_690_review,21.html another review the GTX 690 is getting 80% scaling on BF3,
bit-tech seems to show anything from 70 to 88% scaling or so.
Soldato
All the time I used to run 2 card setups I used to say to myself, What's the problem here, first card renders a frame then the second card renders a frame,
why does it not always give me a big boost ! can not be to hard for those driver wizz kids to sort out....but it never happened, when it did the benchmarks looked good but the games never played to expectations.
Now I have not used 2 cards for a couple of years, but what I see around forums is the same issues, sure maybe when they work well, they are better than than they used to be, but they are not there yet !
The point is You just can not say a card is 95% better "PROVIDING" You only run these games !
This new Titan we should be able to says is X% better across the full board
Yes I can say that because it's true. Your base of comparison is a fallacy. In fact it's a little bit far to even call it a comparison in the first place.
You seemed to have missed the point where I said that the less than perfect scaling is due to software not hardware so I'll say it again....
You can't say that the 690 is only 45% faster than the 680 because that is factually incorrect at a high level. At a low level of comparison of application to application then you will be correct overall but this misses the point.
I understand and agree with your underlying point that a faster single GPU is a better solution overall than a slower multi GPU set up but that isn't what you're saying and you can't use non-scaling software as basis to try over-exaggerate the drawbacks of multi-GPU.
When talking in real terms the 690 is 90-95% faster than a 680. I know this because it's common sense based on the hardware and the fact I have used 680 SLI.
Last edited:
Soldato
- Joined
- 30 Nov 2011
- Posts
- 11,356
All the time I used to run 2 card setups I used to say to myself, What's the problem here, first card renders a frame then the second card renders a frame,
why does it not always give me a big boost ! can not be to hard for those driver wizz kids to sort out....but it never happened, when it did the benchmarks looked good but the games never played to expectations.
Now I have not used 2 cards for a couple of years, but what I see around forums is the same issues, sure maybe when they work well, they are better than than they used to be, but they are not there yet !
The point is You just can not say a card is 95% better "PROVIDING" You only run these games !
This new Titan we should be able to says is X% better across the full board
not really, it's a valid comparison... my 7970 *should* be roughly level with my 670, but in MWO it spikes from 45% usage to 90% resulting in drops in frame rate down to 25-28 FPS
that's a driver problem, it can affect single cards just as much as dual cards, would it be fair to include MWO in a 7970 review and drag the average down to 80% of a 670 because of 1 dodgy game / driver combination?
Soldato
I got near 100% scaling on 680 SLI (306.xx drivers) so no it isn't. I also got close to 100% scaling in 3DM11 and Heaven 3.0 (i.e. applications which take advantage of multi GPU).
Of course you aren't going to get maximum scaling at 1920*1200 - that's so obvious I'm surprised that you've stated it.
The point is more that when talking about "% faster" it's not correct to base it on applications which don't benefit from multi GPU or are CPU limited. If you want to get into a low level comparison then go ahead and compare on an application to application basis but when talking at a high level, the 690 is two slightly downclocked GK104 680 GPUs which directly translates to an almost 95% bump in performance.
I don't disagree in the slightest that depending on what you do with those GPUs may see you achieve a lot less than perfect scaling.
Last edited:
Soldato
Firstly, it's not crossfire.
Secondly, in benchmarks I got near perfect scaling going from one GPU to two and in games like BF3 the same was observed.
Thirdly, it just is around 95% faster - 90% at worst - so if you're surprised this says more about you than it does my percentages.
Yeah that sounds about right. BF3 like BFBC2 will use as many GPUs as you can throw at it.
Soldato
- Joined
- 30 Nov 2011
- Posts
- 11,356
BF3 at 2560x1600, 680 38 FPS, 690 69FPS - 92% scaling (680 SLI 70FPS so statistically the same)
Crysis warhead at 5760, 680 36.5 FPS - 690 64.6 FPS 89% scaling (launch day drivers)
using low resolution examples for SLI scaling is poor journalism
Last edited:
Soldato
The Titan is going to be stupid expensive. Unless your in the market for a 7990 or a 690, I wouldnt be waiting on this.
Yes, so far it's just one rumour from Swe overlockers that everyone keeps repeating over and over.
Permabanned
+1
Don't really care for the Titan. I'll be waiting for a 690 price drop and get one of those