Nvidia still releasing the 7 series this year?

[Dave2150]

You have nothing to back your claims, the same as I don't. The only rumours floating about ar that the 8 series will still be on 28nm and Volcanic islands will be on 20nm.

Whilst you act out your superiority complex and know it all fashion, you have absolutley no evidence of the claims you have just made. You have not provided any informed evidence of you being remotely correct.

http://www.kitguru.net/components/g...adeon-hd-8970-and-8950-launch-plans-revealed/

Going from that slide, you can see that the 8970 is on 28nm (not sure why you keep on about 20nm?).

Whilst I thankyou for allowing me to have an opinion, bring some evidence next time you wish to call me delusional or uninformed

Edit:



Not sure where you are getting your figures but you are massively wrong stating that the 7970GE is ~ 2X the performance of the 6970

When we look at Anandtech, you can see that your guesstimate of 2X the performance is massively delusional.

http://www.anandtech.com/bench/Product/509?vs=618

Just using the 1920x1200 benchs:

7970GE is faster than the 6970 by 50% in Crysis Warhead
7970GE is faster than the 6970 by 49% in Metro2033

Just 2 examples of how wrong that is and the increases are far less with higher resolutions.

Those benches are very old, using old drivers. I guess you'r not well informed that the newer drivers have seen massive performance gains, which is surprising seeing as you imply you know everything.

 

[imginy]

Am I right in thinking that Titan being new it's going to get better?

Maybe only 25% ahead of 7970 now but I would have thought as drivers improve it could easily be 40-50% ahead.

I can't believe how well the graphics card companies can keep secret there release dates and spec of future cards.

 

[Duff-Man]

Am I right in thinking that Titan being new it's going to get better?

Maybe only 25% ahead of 7970 now but I would have thought as drivers improve it could easily be 40-50% ahead.

It may gain a few percent, but it's unlikely to be a lot. While the Titan has a lot of 'new' features related to compute, the core of the architecture that's used in gaming is very similar to the 680, and nvidia have already had a year to optimise drivers. We won't see the level of improvement we've come to expect from a brand new architecture, unfortunately.

 

[imginy]

It may gain a few percent, but it's unlikely to be a lot. While the Titan has a lot of 'new' features related to compute, the core of the architecture that's used in gaming is very similar to the 680, and nvidia have already had a year to optimise drivers. We won't see the level of improvement we've come to expect from a brand new architecture, unfortunately.

Ok thanks for the info

 

[Diagro]

Those benches are very old, using old drivers. I guess you'r not well informed that the newer drivers have seen massive performance gains, which is surprising seeing as you imply you know everything.

Oh he knows, Its just he either neglected to look at the date of the image he linked or he simply used that example for its older driver results.

Anything to justify the Titan purchase

 

[Wh00pS]

Anything to justify the Titan purchase

No one needs to justify anything to anyone, it's thier money and they do with it as they please.

Personally i dont care if the 8970 is slower than Titan, as long as the performance is close and the price is £550 or less then i will buy one and be very happy with my purchase.

 

[yashiro]

Have there been any rumours on if/when Nvidia will drop the FTP currency incentive and go back to providing real games?

 

[spoffle]

Have there been any rumours on if/when Nvidia will drop the FTP currency incentive and go back to providing real games?

I think they've done that because at the moment, they're not really working with many big name developers developing big games.

Their TWIMTBP programme has been stagnant for a while now.

 

[spoffle]

GPUs are parallel computing devices (much unlike the CPU). Given this, performance is largely determined by the number of transistors you can cram into a die.

This leads to large performance jumps when the manufacturing process shrinks; going from 40nm to 28nm allows more than twice as many transistors to be crammed into the same area, and the drop to 20nm will be similar. TSMCs 20nm process is about 18 months away from mass production, so this is when we will see "titan beating" cards.

Inter-process refreshes offer much smaller gains, precisely because they are restricted to using the same size transistors. Yes, the engineers can usually eek out a few percent extra performance by improving the pipeline efficiency, and yes sometimes the clockspeed is bumped up a little. But the improvements are typically small - 15-25% is typically what we have seen from these inter-process refreshes over the past decade.

There is no reason to believe that this time around will be any different. Yes, AMD or Nvidia COULD make a faster GPU than the Titan, on 28nm, by using a massive die size. But the larger dies are much more complex, less power efficient, offer poorer yields, and take longer to "iron out the bugs". AMD have not gone for the "big die" approach since the 2900xt, and have shown no intention of doing so in the near future. Nvidia already has their "big chip" - the GK110 on which the Titan is based. The only reason that the Titan exists is due to the Tesla K20 (a card designed for scientific computing). Nvidia are not replacing the K20 until Maxwell, so there will be no "Titan buster" coming from them either.

So no, it isn't "delusional" or "wishful thinking" as you suggest, it's simply a case of cold hard logic. Historically, technically and realistically there will be no "Titan busters" on the market until 20nm cards arrive, and that isn't going to be for *at least* 12 months, realistically more like 18. The high-end 28nm refresh cards will get close to Titan performance, but it's highly unlikely to be "beaten" in terms of overall performance until 20nm rolls in.

Aren't there rumours that Globalfoundries might be producing the 20nm GPUs for AMD?

I was basing my post and comments about delusional wishful thinking on 20nm GPUs.

But in addition to that, with this being the start of a new console generation, as in, the games are largely already in development, games requirements will increase, the reason why AMD has left the 7 series on top for so long is because there's little benefit in bringing out anything faster because for the most part, most games don't need it yet, and by that, I'm talking for those in the mainstream of playing games at 1080P rather than niches of 2560x1440/1600, 5760x1080/1200 or even 7680x1440/1600.

 

[psychas]

Personally i dont care if the 8970 is slower than Titan, as long as the performance is close and the price is £300 or less then i will buy one and be very happy with my purchase.

bring back the pricing like it was 6970 days , bough it on release day was cheaper then 300£

 

[Duff-Man]

Aren't there rumours that Globalfoundries might be producing the 20nm GPUs for AMD?

I was basing my post and comments about delusional wishful thinking on 20nm GPUs.

But in addition to that, with this being the start of a new console generation, as in, the games are largely already in development, games requirements will increase, the reason why AMD has left the 7 series on top for so long is because there's little benefit in bringing out anything faster because for the most part, most games don't need it yet, and by that, I'm talking for those in the mainstream of playing games at 1080P rather than niches of 2560x1440/1600, 5760x1080/1200 or even 7680x1440/1600.

There have been rumours (well more than just rumours...) that AMD will switch to global foundries as its primary manufacturing for GPUs for a while now. The same theories were in place for the 28nm generation.

It's fairly clear that AMD would love to have a viable alternative to TSMC - and frankly I hope that GloFo does "come good" at 20nm, as having a second active manufacturing source would be great for the consumer. I'm a little skeptical that it will come to pass this time around (GPUs are incredibly complex and require a very stable process in order to achieve high yields), but I don't know enough about the specifics of the plant to make any informed statements. I hope it happens, but I'm not holding my breath that 20nm will be stable at GloFo before TSMC.

The argument that "more GPU power is not needed" has been brewing for a while. Certainly with games designed for current-gen consoles, the major benefit of a super high-powered GPU has been restricted to those using high resolutions orm multiple screens - this has arguably been the case for a few years now. It doesn't seem to be slowing demand for ever-faster GPUs though... Consumers in these high-end "luxury" markets tend to buy from a position of 'want' rather than 'need'.

The new consoles, and the resulting multi-platform games, will certainly refresh demand for high-end GPUs, but I'm not sure that's the main driving force behind delays. Bringing in a GPU refresh is an expensive process, and I think that both AMD and Nvidia are content to keep bringing in money from their current lineup for the time being. Also, delays to 20nm mean that the "true" next-gen cards are being pushed further back, and delaying the 28nm refresh may be partially a result of this (i.e. will allow a reasonable lifespan for the refresh cards even if they're introduced late this year).


Getting back to the original point - I agree that the 20nm cards will beat the Titan by a fair margin. The real question is when they will appear. 2 years seems very pessimistic, and 1 year seems overly optimistic. Beyond that I can't really say... it would be great to see AMD push out 20nm cards from GloFo at an earlier date, but I wouldn't count on it.

 

[Bacon?]

I think some people are looking at the generation increase in performance in a wierd way.

For example:

4** series to 5** series wasn't a massive increase in performance. The 460 and 560 were roughly equal and the 480 wasnt far behind the 580.

Same with the 5*** to the 6***. I wouldn't be surprised if this next gen was less than 30% performance. It seems its every other gen when you get good performance increase. Ie this gen from last gen.

 

[Rroff]

Getting back to the original point - I agree that the 20nm cards will beat the Titan by a fair margin. The real question is when they will appear. 2 years seems very pessimistic, and 1 year seems overly optimistic. Beyond that I can't really say... it would be great to see AMD push out 20nm cards from GloFo at an earlier date, but I wouldn't count on it.

TSMC is claiming getting the design kit out early will mean they can go into production earlier but theres no indication of realworld timescales. Also claiming production runs will be commencing late November, if everything went according to plan then 12 months from now seems not unlikely but historically things haven't tended to go to plan.

 

[Orangey]

It's been years now hearing about TSMC being rubbish and GF cranking out GPUs "soon". I think Charlie was still working at El Reg when AMD originally spun off its fab business.

 

[Rroff]

I think some people are looking at the generation increase in performance in a wierd way.

For example:

4** series to 5** series wasn't a massive increase in performance. The 460 and 560 were roughly equal and the 480 wasnt far behind the 580.

Same with the 5*** to the 6***. I wouldn't be surprised if this next gen was less than 30% performance. It seems its every other gen when you get good performance increase. Ie this gen from last gen.

Its a bit more complicated than that tho, 400 to 500 series was a simple refresh and you rarely get more than 30% tops from that unless it involves a die shrink but a major architecture change and a die shrink usually brings a big performance change. Then you get complications like the 5000 to 6000 series where the architecture changed a lot but it was on the same process size.

 

[Duff-Man]

I think some people are looking at the generation increase in performance in a wierd way.

For example:

4** series to 5** series wasn't a massive increase in performance. The 460 and 560 were roughly equal and the 480 wasnt far behind the 580.

Same with the 5*** to the 6***. I wouldn't be surprised if this next gen was less than 30% performance. It seems its every other gen when you get good performance increase. Ie this gen from last gen.

Yes - when the manufacturing process shrinks, it's possible to cram many many more transistors into a given space, leading to big gains in performance (typically close to 100%). Within the same process gains are much smaller, as gains rely on improving efficiency and/or increasing chip size (so typically around 20%). In recent years we've generally seen a process shrink every other generation.

... see earlier posts in this thread for more detail.