AMD: Southern Islands (7*** series) Speculation/ Rumour Thread

Given that in the UK 680s are a couple of quid more than 7970, I doubt we'll see anything of these price drops. AMD and NV are screwing us over BIG time. They are already £40-50 more than they should be here, even taking into account VAT. These companies need to stop taking the piss in Europe.
 
Given that in the UK 680s are a couple of quid more than 7970, I doubt we'll see anything of these price drops. AMD and NV are screwing us over BIG time. They are already £40-50 more than they should be here, even taking into account VAT. These companies need to stop taking the piss in Europe.

Actually going from this thread http://forums.overclockers.co.uk/showthread.php?t=18392298 7970's will be around £360 and 7950's will be around £299 in a few days.
 
I think a lot of people (including myself) are just skipping this generation. The price is just too much for what is a fairly incremental upgrade from a 5xxx/6xxx series card, especially if you're on a single HD monitor. By the time the prices come down, the refresh will be on the horizon.
 

Interesting, if true. And I have no doubts it probably is true.

AMD will have made it ~4 weeks without having to reduce price. 479 USD is certainly lower than I was expecting but it makes sense for where the stock clocked 7970 is in relation to the GTX 680 when both aren't OC'd.

This also seems to imply potentially good news for people wanting a GTX 680 as it may be indicating that the supply versus demand situation for the GTX 680 is improving. I don't think GTX 680 will be dropping as I foresee it still being somewhat difficult to obtain in the US for at least another month or so. Although it might possibly drop in price in Europe in response to AMD's actions as supply versus demand is much better in Europe. Or they'll just allocate more European stock to the US to keep the current price point there.

Regards,
SB
 
NCIX is now reflecting the price drops. Reference Radeon 7970's are mostly $479, xfx black edition is $499, some of the other OC editions haven't moved yet.
 
So there is no more significant performance gains to be found in the drivers after all and competition has to be done on price now?

and what if both amd and NV both have large yet proportional driver increases. Wont most of the seen performance increases(VLIW to GCN) for AMD not come from better drivers but better shader utilization for things like compute shaders which willwill largely be out of the drivers control?
 
and what if both amd and NV both have large yet proportional driver increases. Wont most of the seen performance increases(VLIW to GCN) for AMD not come from better drivers but better shader utilization for things like compute shaders which willwill largely be out of the drivers control?
I was refering to this:

http://forum.beyond3d.com/showpost.php?p=1607915&postcount=1848

and wondered if everything that was to be learned is already extracted and only game specific optimisations are left.

Driver improvements have little to nothing to do with pricing surely?
Usualy not since they don't account for much. But with two cards going almost head to head, it could be enough to gain the upper hand in higher resolutions and together with better compute performance AMD could easily keep asking 549$ for Tahiti.
 
Usualy not since they don't account for much. But with two cards going almost head to head, it could be enough to gain the upper hand in higher resolutions and together with better compute performance AMD could easily keep asking 549$ for Tahiti.

How many times should it be repeated that NV has the upper hand in the discrete market with 60 or more % market share?
How many times should it be repeated that AMD NEEDS lower prices in order to sustain interest in their products, when they are overpriced, significantly less people would tend to go there.
How many people/ gamers do actually care about that compute performance?
What exactly is the performance difference between Pitcairn (costing around 300-350$) and this same Tahiti in compute? Is it even worth to pay so much more for marginal improvement in most compute scenarios?

And the last but not least. IF YOU CARE so much about the company, you are always free to donate. But don't give wrong ideas, and don't protect and justify the interest of this rich corporation. Please! :rolleyes:
 
Take a deep breath and count to ten.
.
.
.
.
.
.
.
.
.
.
Finished? Then read on.

I wasn't protecting anyone's interest, just concluding AMD doesn't have anything up its sleeve and all is left is to drop prices.

So what if nV has the majority of discrete? AMD still has the majority of DX11 marketshare, so what? My whole point was, they wouln't NEED to lower prices if they had something comming. Actualy a lot of people do care about compute performance, given how CUDA was touted as a nVidia advantage. And Tahiti is a lot faster in double precision than Pitcairn, duh. But it doesn't matter if anyone cares or how fast it actualy is. It is a checkbox feature that AMD could ask the money for.
 
Last edited by a moderator:
If you're interested, here's some GPGPU benches on GTX580, 680 and HD6970, 7970. It's in finnish but the graphs are quite universal, as long as you remember "suurempi on parempi" means bigger bar is better, while "pienempi on parempi" means smalle bar is better :)

edit:
Oh ye, the link :D http://muropaketti.com/artikkelit/naytonohjaimet/gpgpu-suorituskyky-amd-vs-nvidia,2

Thanks. Does the article explain why MilkyWay is so much faster on AMD Hardware, and so slow on Kepler vs. Fermi? Does it use DP?
 
Thanks. Does the article explain why MilkyWay is so much faster on AMD Hardware, and so slow on Kepler vs. Fermi? Does it use DP?


I was in the same question about this, but i dont know this one .

Maybe not optimised at all .

If we remove the 7970, what is the most incredible ( who is sometimes even 10x faster ) , is the lost of performance in compute against the GTX580 ... ( personnally i will have use a GTX480, the 580 was allready slower of it )..


Last quarter of 2011, Not the best.... Its funny enough as AMD is now leader in global GPU market... ( under Intel ofc ) ( The lost in discrete gpu market is essentially due to 2 factors on last quarter 2011, essentially cause AMD have move low end class gpu to the CPU, but the fact AMD was ready to launch a new series is ofc a good part of it too ... )... Its start to be really complicated with the move on CPU of gpu low end class, to compare both....

You know thoses numbers are essentially a report of Low class gpu, as middle and high end dont represent both, more of 15% of the total market of gpu ( by brand ).

For give you a funny example, the sales of the GTX560 ( all version include TI and non TI ) have doubled since September-October 2011... passing from 4% to 8% ...

For be honest, i dont count too much on gpu number sales on this first quarter: 7970-7950 have been launch in january, 680 is the only card from Nvidia.. and all cards are not cheap at all. ( including 7870-7850 )... the availability of both 7970 and 680 on their respective launch have been far of enough and the market is quite discorded. We was all wait for a massive 2012 Tsunami of cards, new series and for both AMD and Nvidia, this is not the case.. the card appears slowly on the market ( with AMD who have at least now a panel of cards: 7750-7770-7850-7870-7950-7970 vs only a the GTX680 from Nvidia )..
Its excellent to see the GTX560 sales a lot at his respective price, or this is by far not the best card yet lol, at least performance wise. ( 7870 is on 580 performance level )
 
Last edited by a moderator:
Thanks. Does the article explain why MilkyWay is so much faster on AMD Hardware, and so slow on Kepler vs. Fermi? Does it use DP?

No it doesn't, the article isn't very in-depth so to say, and the reviewer really has no expertise in gpgpu's, there's just basic explanations of what each app does, and then the benches themselves.
 
Back
Top