"Polaris will support hardware 4K h.265 encoding and decoding at 60 FPS,
DisplayPort 1.3, and, at long last, HDMI 2.0a. The latter was missing
from AMD's recent Fury and 300-series of GPUs, which instead featured
HDMI 1.4a that limited 4K signals to 30 FPS at 60Hz, making them less
than ideal for use in the living room with 4K TVs."
1 Like
Sweet! Cant wait to sell my 390s for 490s....or maybe a Fuxy X2 (or X squared?).
Hopefully this means that AMD will start to lead the GPU industry in the embedded market.
This could be a really good thing for AMD; But why such an early release? From my standpoint that's only going to hurt sales totals because a lot of users have just recently purchased AMD's 300 series or Nvidia's 900 series. Needless to say though I hope this will be a win for AMD, as well as Zen.
This is just an announcement of the tech, it's not going to release for at least another 6 months. By then the Fury X will be a year old and ready to be phased out.
I'll sell my Fury in 6 months lol
2 Likes
That's lovely and all, but how big with the die be?
Their working sample is estimated to be 120mm², and that's at the low end. If this follows the old way of the Radeon, the flagship won't be larger than 380mm². AMD never wanted to get rid of the small-die strategy, it just had to happen to continue gains on 28nm. Now that we're out of that rutt the die sizes can be reduced.
The glory of Gaben smiles upon AMD
1 Like
Are you just memes now? You seem to be memes.
Shit, not sure if he's on to me or if he's just joking.
That's a good question...moving from 28nm to 14nm along with HBM could decrease the overall form factor of the card while still creating a substantial performance gain or enable multi-gpus on the same standard size card....time will tell.
Seriously though it's about time the process got smaller, how long has it been 28nm? 5 years? Would be interesting to see if this will improve overclocking as the overclocks on recent flagships have been sad.
And obviously all of us are hoping for performance gains in general
We need huge dies tho, none of that sub 400 peasantry.
wonder how many hertz it'll be capable of since 14nm LPP is for low power.
We absolutely do not need big dies. Big dies means less production volume, higher transistor variance, lower yields, and above all higher consumer pricing. Big dies also take more time to tape out, more time to manufacture, and often require higher PCB complexity.
Good thing to point out. If we take Anandtech's article at face value, than RTG is planning to run both 16nm and 14nm. I would assume that their more complex designs will be 16nm, and 14nm will be left to the low-power, high-efficiency designs populating the sub-$200 price point and mobile markets. 14nm may be <1GHz exclusive while 16nm may see 1100+.
Hmm interesting, i might wait with upgrading my gpu for this upcomming launch.
I am excited. Hopefully the will decrease the prices of current gen second hand cards into my price category ;)
Low TDP gives me a hard-on.
Was that too blunt?
1 Like