AMD Pirate Islands Radeon R9 300 Series
The colleagues over at WccfTech posted something about the next generation Radeon graphics cards. They dug up an old (what is claimed to be) roadmap which shows details about three GPUs, the R9 390X, R9 380X and R9 370X. There is no validation regarding the source material and their post is highly speculative. Then again .. rumors always start somewhere and props to them for retrieving that info.
Now in their news item posted they already claim themselves that all information is to be taken with a pinch of salt, which really is the proper way to address this information.
The source leak indicates future Pirate Islands GPU code names and some specs. Specs NEVER are final which makes me frown up-on the posted material, anyway here's what they are claiming:
--
According to these details all GPUs will be based on the TSMC 20nm Node and will be true Pirate Island cores, which is why nomenclature of the dies will be derivative in nature. Another thing mentioned is that all Pirate Island GPUs will have the DirectX 12 Hardware Feature Set. According to these details the R9 370X is slated for announcement/arrival sometime this July-August this year. The R9 370X will feature a ‘Treasure Island XTX’ core and supposedly has 1536 Stream-processors, 96 Texture Units and 48 ROPs on the uncut die. Of course this road-map precedes the report we got of TSMC having a little trouble with 20nm so I am not sure how valid the time frame is anymore.
Now according to the details, we will see the R9 390X before the R9 380X. The R9 390X will feature the ‘Bermuda XTX’ core and feature a staggeringly high number of SPs, ie. 4224. The Texture Unit count is 264 with 96 ROPs. It is slated to show up late 2014 around November or so. Once again keep in mind the TSMC comment above. Finally the R9 380X which will feature the ‘Fiji XTX’ core and the following die configuration: 3072 SPs, 192 Texture Units and 72 ROPs and should appear sometime early 2015. Now interestingly the R9 390X ‘Bermuda XTX’ GPU is slated to have 512 Bits of Memory Bandwidth while as the R9 380X is slated to have 384 Bits with the R9 370X coming last with 256 Bits. Now the road-map even hints at the clock speeds of AMD’s future GPUs but I think it is prudent to mention that all these tech specs at this point in time are about as solid as the wind. However let us continue, with the help of this table:
GPU | Codename | Core | Stream Processors* | Texture Units* | ROPs* | Core Clock* | Memory Clock* | Memory Bandwidth* |
R9 370X | Treasure Island | Treasure Island XTX | 1536 | 96 | 48 | ~900Mhz | ~5Ghz | 256Bit |
R9 380X | Fiji | Fiji XTX | 3072 | 192 | 72 | ~900Mhz | ~6Ghz | 384Bit |
R9 390X | Bermuda | Bermuda XTX | 4224 | 264 | 96 | ~1000Mhz | ~7Ghz | 512Bit |
* Values are not indicative of the final product.
Posts: 22075
Joined: 2008-07-14
Just rumors, but either way doesn't affect what AMD does...unless TSMC can't pull off 20nm for some reason.
Senior Member
Posts: 9312
Joined: 2008-01-06
If that R9 390X spec is true... HOLY DAMN!!!
Think I will hold off getting a second 7970 til I hear more about that potential beast!
Senior Member
Posts: 1230
Joined: 2012-05-22
Wasn't Nvidia 20nm delayed till 2015?
Speculation based at least partially on fact...
Nvidia will wait till AMD releases Pirate Island first then figure out how little or how hard they will have to try to beat it.
Speculation based on the baseless idea that these companies actually have the ability to make abrupt and substantial architectural changes to a chip regardless of the fact that they're typically already in final production and would have to kill things full-stop, make changes, go through the tape-out process, obtain re-certification and approval, then re-start the entire pre production/ramping up/full production process all over again.
The only thing that can be done so far along is increase mem/core clocks if the silicon process is mature, with good enough yields, to allow for the additional voltage while avoiding leakage and other heat inducing factors.
Just rumors, but either way doesn't affect what AMD does...unless TSMC can't pull off 20nm for some reason.
The possible delay of nVidia's Maxwell has a direct correlation to when AMD will be able to release their chips because they're both being manufactured at TSMC and right now it seems TSMC's 20nm wafer capacity is completely booked till the end of Q3 2014 by smartphone and tablet manufacturers.
Just the facts Ma'am...
TSMC capacity booked to the end of 3Q14
Cage Chao, Taipei; Steve Shen, DIGITIMES
The production capacity of Taiwan Semiconductor Manufacturing Company (TSMC) has been booked through the end of the third quarter of 2014 thanks to orders from handset solution and PC chip vendors, according to industry sources.
The planned roll-outs of new smartphones by brand handset vendors and the upcoming 4G services in China have pushed chipset solution vendors to place wafer start orders at TSMC, the sources noted.
Consumer IC suppliers are also placing short lead-time orders at TSMC as demand from downstream and end-market device makers is likely to pick up soon, added the sources.
As a result, TSMC's 8- and 12-inch fabs are currently operating at 100% and over 90% of their capacity, respectively, said the sources, adding that the foundry's 20nm and 28nm production lines are fully occupied by orders for smartphone and tablet solutions.
http://www.digitimes.com/news/a20140402PD212.html
Senior Member
Posts: 22075
Joined: 2008-07-14
AMD isn't limited to just TSMC. lol
AMD also has a deal with GloFo for GPUs... So, if TSMC can't put out GPUs for them, AMD can simply fall back to GloFo.
Posts: 31495
Joined: 2005-01-08
Yeah, I think AMD should do that, it's only wise not to go first.