Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

AMD Confirms RX 480 At $199 USD, Other APU & Polaris Announcements

SoyCapitanSoyCapitan said:
Zen-Vega will come to nMP.

Polaris to MBP and iMac. Apple will have seen mobile Pascal by now but the evidence is in AMD's favour ATM.
Click to expand...

Zen is a CPU. Vega is probably going to have too much heat for nMP. We'll have to pay close attention to AMD's rollout of Polaris Fire series cards tomorrow to see if that card fits the nMP.
 
xsmi123 said:
Zen is a CPU. Vega is probably going to have too much heat for nMP. We'll have to pay close attention to AMD's rollout of Polaris Fire series cards tomorrow to see if that card fits the nMP.
Click to expand...
...and you know that Apple has been fully briefed on the GTX 1070M and GTX 1080M - perfect for Imac and nMP....
 
SC, hope you're wrong about Zen on the nMP. I'd rather stick to Xeons.
I'm guessing whatever comes out tomorrow regarding FirePros it's what we'll be seeing in the nMP.
Maybe even a hint at Vega?
[doublepost=1469349132][/doublepost]The new rMBPs should get either Polaris 11 or cut down Polaris 10. Even Apple wouldn't step back from the current GPU. Most models use iGPUs anyway, exception being the top 15".
Too bad Kaby Lake probably won't make it there as well.
 
FUDzilla was correct at number of ocasions lately, and they are much, much more credible than for example WCCFTech.
 
That is surprising. Nvidia is pushing hard the releases, even those that were supposed 10 nm and 2018. What is funnier, one of sources on the internet claims that Nvidia pays 125$ for each GTX 980 Ti sold for lower price, 100$ for GTX 980 and 50 for each GTX 970 to retailers. That is the effect for pushing the release of all pascal cards.

What the hell is going on? Why are they rushing so hard?
 
Yeah, it's not like they actually need to do so.
Either trying to kill off AMD once and for all or just for the bragging rights.
That might blow in their faces though.
 
It looks like it can be confirmed that Vega GPUs will be produced on 16 nm TSMC process.

Very interesting.

Also sources look at:
http://mobile.reuters.com/article/idUSKBN0U500P20151222
http://www.timesunion.com/tuplus-bu...s-wins-chip-contract-with-Defense-7966717.php
http://www.forbes.com/sites/patrick...14nm-manufacturing-with-samsung/#3150280d50fb

And biggest one, however behind paywall:
http://semiaccurate.com/2013/07/12/apple-has-their-own-fab/
http://semiaccurate.com/2014/04/17/semiaccurate-right-apples-foundry-plans/
http://semiaccurate.com/2013/12/17/apple-samsung-intel-foundry-plans/
http://semiaccurate.com/2013/08/26/a-third-player-emerges-apples-foundry-plans/
http://semiaccurate.com/2013/03/25/what-is-apple-doing-at-14nm/
IBM signs partnership with Apple. IBM sells the fab and technology, and engineers to GloFo(FD-SOI). Samsung settles with Apple all patent disputes. Samsung synchs with GloFo their technology when it goes for manufacturing process. AMD needs to feed BIG customer with GPUs.

Looks like huge update for Apple computers incoming. 5 fabs total will produce chips for AMD soon. 2 Samsung, 3 GloFo.
 
Last edited:
To be fair. It comes from coupling "words" coming from silicon industry with the news about AMD saying that TSMC and GloFo are their fab partners for 14 nm GPUs and they have the capacity to use the Samsung in addition to them.
AMD has strong foundry partnerships and our primary manufacturing partners are GLOBALFOUNDRIES and TSMC. We have run some product at Samsung and we have the option of enabling production with Samsung if needed as part of the strategic collaboration agreement they have with GLOBALFOUNDRIES to deliver 14nm FinFET process technology capacity
Click to expand...

And Samsung is getting some love from AMD, also, as it appears to be. I do not know what it means, nor how Vega will perform.

One small, little thing:
https://radeonopencompute.github.io/index.html - Another AMD initiative for HPC.
And this looks very interesting.
 
ActionableMango said:
What the heck is Apple making in this fab? I only ever read about them using other fabs.
Click to expand...
Actually, it is all about IBM technology(FD-SOI), knowledge(Engineers), and capacity. And selling the whole Fab to GloFo. Apple wanted good GPUs for their computers.

AMD have had WSA(Wafer Supply Agreement) with GloFo, but their technology was rubbish compared to TSMC, and it appears it might still be. There is however this FD-SOI technology that allows for scaling down the transistors while lowering the leakage, and maintaining the very high clocks. GloFo has right now the technology, knowledge, etc to bring it to market.
AMD has fab and technology(Polaris) that can be pipe cleaner of WSA. Apple can consume most of deliveries of Polaris chips.
IBM is happy because they do not need to feed the fabs.
GloFo is happy because they can grow in technology, revenue, and mindshare.
AMD is happy because they can get more revenue, do not loose money on WSA, and whats more - get some money from WSA.

For me it looks like quite good business for all of the parties. And the news about Samsung producing chips for AMD, because of the needs of some partner(Polaris chips are produced and designed on the same process for both Samsung and GloFo) adds to whole picture.

Now lets wait for what is about to happen.
 
  • Like
Reactions: ActionableMango
koyoot said:
To be fair. It comes from coupling "words" coming from silicon industry with the news about AMD saying that TSMC and GloFo are their fab partners for 14 nm GPUs and they have the capacity to use the Samsung in addition to them.


And Samsung is getting some love from AMD, also, as it appears to be. I do not know what it means, nor how Vega will perform.

One small, little thing:
https://radeonopencompute.github.io/index.html - Another AMD initiative for HPC.
And this looks very interesting.
Click to expand...

I don't think its any surprise that AMD will be producing Vega at TSMC. They have basically been the sole manufacturer of GPUs for quite a bit. Nvidia is already making GPUs up to 600 mm2 on TSMC's 16 nm process. It was surprising when AMD chose Global Foundries for Polaris and I think its partially to blame why Polaris 10 is less competitive in efficiency.
 
Stacc said:
I don't think its any surprise that AMD will be producing Vega at TSMC. They have basically been the sole manufacturer of GPUs for quite a bit. Nvidia is already making GPUs up to 600 mm2 on TSMC's 16 nm process. It was surprising when AMD chose Global Foundries for Polaris and I think its partially to blame why Polaris 10 is less competitive in efficiency.
Click to expand...
Both have exactly 36 GFLOPs/watt, and similar performance in DX12, and Vulkan, per watt. Only DX11 with high overhead on AMD GPUs is where AMD's efficiency tanks. And drives people's mindshare...
 
Radeon Pro WX7100 looks great on the nMP.
But they could have taken the mem to 8G...

Will Apple use SSG? The nMP already has something of the sort, although not to this end.
 
Raja have said that there will be one more GPU, soon ;).

So I think we have the GPUs for Mac Pro 2016:
Radeon Pro DX300 - 2048 GCN cores, 4 GB of RAM, 256 bit memory bus,
Radeon Pro DX500 - 2304 GCN cores, 8 GB of RAM, 256 bit memory bus,
Radeon Pro DX700 - ? GCN cores, 4/8 GB HBM1/2, 4096/2048 bit memory bus with 512 GB/s.
 
  • Like
Reactions: ManuelGomes
Who says that Apple have to use two SAME GPU's in next nMP? Let's say that the GPU is a DX series Polaris 10 Pro and XT. But for GPGPU they use Fury Nano. That way they'd get best of both worlds: latest GPU tech with DP 1.4 and fastest GPGPU AMD has at the moment for under 175W TDP class.
 
Last edited:
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back