AMD: Speculation, Rumors, and Discussion (Archive)

Status
Not open for further replies.
Dont really buy on this 199$ .. dont know exactly, what model we are talking about there.
Superficially, it doesn't seem like Polaris 11 would fit that price or performance range.
For Polaris 10, if it's not the highest SKU, then it's a potentially pleasant surprise on the performance front for the X SKU.
If it is the top SKU, the price is rather nice. I would wonder if AMD would be going by the old joke about losing money for every unit sold, but making it up in volume.
 
Pricing needs to be one part of AMDs strategy here. Would also like to see their strategy to counter nVidia's distribution channel and brand advantage.
 
$200 for a <150W card with the performance of a R9 290 would be a steal. And I mean HD4850 levels of steal.
However, I wouldn't put too much hope on the $200 version having 8GB. Though 4GB should be more than enough for 1440p in the foreseeable future.

Maybe it's $200 for a 4GB Polaris 10 with a small PCB, lower clocks and 4GB with the same approx. performance as a R9 290 and GTX970 at about 4.5TFLOPs (currently considered VR-minimum) and then a larger 8GB Polaris with higher clocks (5.5 TFLOPs) that matches the R9 390X in real-life performance.



Regardless, if these prices are confirmed, we may be seeing AMD trying to redo the RV770's success, which gained them a good chunk of marketshare at the time (and consequently brand recognition).


EDIT: Got rekt, changed.
 
Last edited by a moderator:
Hmm this seems more like the rv670 series though, 2 cards needed to get to a 1080 gtx performance, (if those leaked 3dmarks are correct), so its not like its in 10 to 15% of the top end card so far released.
 
AMD's partners apparently think Polaris will double their market share: http://www.hardware.fr/news/14651/computex-polaris-doubler-pdm-amd.html
I think increasing to 30% is fairly realistic, doubling it only in midrange/low end segment, no that doesn't sound likely. That means they would be trying to take marketshare away from not just nV's midrange *which isn't out yet* but also take away from performance segment too, I don't see that happening with these cards.
 
A rumor(from videocardz.com editor) to help place the RX480 performance: 2 cards in CF faster than a 1080.
If true the 199 dollars price you could have for 400 dollars a system faster than a 1080.

mGPU is a crapshoot. Until that changes single GPU is the way to go.
 
RV770?
RV670 was 3870/3850.

Ah, the RV770. That one cleaned up nice for a die shot.
AMD still uses that shot to this day, even the FinFET GPU marketing.

There have been other AMD GPU die shots that we've pored over since then, but for whatever reason that one for me seems like the right mix aesthetically. It wasn't too packed to lose recognizable features at a coarse level, and it was in that exclusive club of chips that AMD decided to show off.
 
I think increasing to 30% is fairly realistic, doubling it only in midrange/low end segment, no that doesn't sound likely. That means they would be trying to take marketshare away from not just nV's midrange *which isn't out yet* but also take away from performance segment too, I don't see that happening with these cards.

~7.5million >$350 GPUs sold in the last ~1.5years (GTX970 release).
Just like Nvidia did with GTX950, they are targeting the tens of millions of people that play WOW/DOTA/LOL and are playing with <$300 GPUs that are >2-3years old while giving them the benefit of solid DX12 and VR performance for future games.

Doubling discrete desktop marketshare by the end of the year isn't outside the realm of possibility but obviously depends on when Nvidia is able to release their counterparts.
 
Last edited:
~7.5million >$350 GPUs sold in the last ~1.5years (GTX970 release).
Just like Nvidia did with GTX950, they are targeting the tens of millions of people that play WOW/DOTA/LOL and are playing with <$300 GPUs that are >2-3years old while giving them the benefit of solid DX12 and VR performance for future games.

Doubling discrete desktop marketshare by the end of the year isn't outside the realm of possibility but obviously depends on when Nvidia is able to release their counterparts.


You have to understand nV took marketshare away from AMD, when AMD wasn't able to compete from 40% to 20% in over a year. How do you think AMD will get back to 40% faster when they will have competition ?
 
Doubling discrete desktop marketshare by the end of the year isn't outside the realm of possibility but obviously depends on when Nvidia is able to release their counterparts.
I think it's unlikely since "price cuts" on both sides would happen way before it got to that point ... and think Nvidia would have more leverage for accepting lower margins without incurring negative returns.
 
Status
Not open for further replies.
Back
Top