videocardzR9 390X Specs Allegedly Leaked

[f1sherman]

Qualcomm has specifically stated they will have no 20 products until next year

they did?
oh well... this pretty much seals the 20nm story

 

[monstercameron]

they did?
oh well... this pretty much seals the 20nm story

yes and...http://www.qualcomm.com/media/relea...s-first-commercial-20-nm-lte-advanced-chipset

 

[ShintaiDK]

Thats just a radio.

Qualcomm will make the first MPU in Q1 2015. MediaTek in H1 2015.

Apple will be the only one with 20nm MPUs in 2014.

 

[f1sherman]

yes and...http://www.qualcomm.com/media/relea...s-first-commercial-20-nm-lte-advanced-chipset

wth is with modems that they always get the new node FIRST D:

 

[tviceman]

Only SITH deal in absolutes! Roadmaps have a tendency to change...

While true (and nice quote!), from my observations, whenever roadmaps change it is almost always not for the better.

 

[tviceman]

wth is with modems that they always get the new node FIRST D:

Modems are usually very small chips and much more evolutionary in design than gpu archtectures.

 

[NostaSeronx]

Did some checking up apparently the R7 M260 influx in India being called "Topaz XT". In which they all turned out to be the R7 M260, Opal Pro with 2 GB DDR3.

The actual Topaz parts will have 4 GB DDR3 using any of;
H5TQ4G63CFR-N2C
H5TC4G63CFR-N0C
H5TC4G63CFR-11C

Tonga - 20-nm - 2 GB HBM + 4 GB GDDR5, August
Iceland - 20-nm - 4 GB DDR3, October (Mostly in Broadwell laptops)

--
6604 Opal XT [Radeon R7 M265]
6605 Opal PRO [Radeon R7 M260]
6900 Topaz XT [Radeon R7 M260/M265]

It must be super confusing for Toshiba.

 

[tviceman]

You heard it here first guys: AMD beats apple to 20nm, pays premium for early limited wafers, also includes first time HBM on unproven node process.

In unrelated news, I saw a unicorn earlier.

 

[NostaSeronx]

You heard it here first guys: AMD beats apple to 20nm

Apple does not make GPUs.

,pays premium for early limited wafers,

No, GlobalFoundries will be in Volume Production by the time these GPUs will be out. TSMC is already in volume while most of the reasons why SoCs aren't coming out. It does not fit in the time table of early introduction.

also includes first time HBM on unproven node process.

20-nm LPM has been the planned node since 2012 after 28-nm SHP and 28-nm HP.

 

[Techhog]

Apple does not make GPUs.No, GlobalFoundries will be in Volume Production by the time these GPUs will be out. TSMC is already in volume while most of the reasons why SoCs aren't coming out. It does not fit in the time table of early introduction.20-nm LPM has been the planned node since 2012 after 28-nm SHP and 28-nm HP.

Unless you have a source that specifically says "20nm AMD GPUs are coming in 2014," stop posting. You seem delusional at this point, incapable of distinguishing between facts and your own educated guesses.

 

[Lepton87]

You seem delusional at this point, incapable of distinguishing between facts and your own educated guesses.

That's the impression that I always get when I'm reading his posts.

 

[NostaSeronx]

Can you provide a road map or some granularity regarding process node
transitions in 2014 and '15 as it relates to thin set by Graphics and APU? Thanks

We are fully top to bottom in 28-nanometer now across all of our products, and we are transitioning to both
20-nanometer and to FinFET over the next couple of quarters, in terms of designs. So we'll continue to do that across our foundry partners.

Here you go.

2014 - Fast Portability aka GPU IP 20-nm (SOC/LPM)
2015 - Fast Portability aka GPU IP 16/14-nm (FF+/LPP)

28-nm : 9T/12T
20-nm : 9T/12T
16/14-nm : 7T/9T

 

[Techhog]

Here you go.

2014 - Fast Portability aka GPU IP 20-nm
2015 - Fast Portability aka GPU IP 16/14-nm

Again, it's saying that they're working on 20nm without a specific timeframe for retail launches. From this, it's safe to say that 20nm is targeted for either late 2014 or H1 2015. There's a possibility of it coming this year from reading this, but it's not a fact. Adding 16/14nm to the mix just comes straight out of left field.

On top of that, AMD's statement about no 20nm this year came after this, so this statement is no longer relevant.

 

[NostaSeronx]

Can you guys talk about the puts and takes of either
moving, or perhaps not moving your next-generation follow-on to Kaveri to the next node here? And how this might affect your decisions to move
other products and families to the next node, or not to move them? Thank you.

So what we've said in the past is certainly this year, all of our products are in 28-nanometer across
both graphics, client, and our semi-custom business. We are actively in the design phase for 20-nanometer, and that will come to production. And
then clearly, we'll go to FinFET, so that would be the progression of it.

Nothing is limiting from the Q1 2014 Analyst Call preventing 20-nm this year. While the second time she does format it weirdly. It does not appear to be a correction but more of an erroneous recollection. Of the previous statement that I have posted.

From the same call after the one above:

Great, and when will we expect you to introduce GPU products at a node below 28-nanometers? Will you have any 20-nanometer GPUs this year
or next year?

David, yes, I think what I said earlier, what we're doing in terms of technology, strategy, we are 28 this year. We have 20-nanometer in design, and
the thin-fed thereafter, so that's the overall product portfolio.

It should be noted that she is clearly talking about the whole portfolio of AMD's products. Not just consumer APUs or consumer GPUs.

 

[Techhog]

Nothing is limiting from the Q1 2014 Analyst Call preventing 20-nm this year. While the second time she does format it weirdly. It does not appear to be a correction but more of an erroneous recollection. Of the previous statement that I have posted.

From the same call after the one above:

"So what we've said in the past is certainly this year, all of our products are in 28-nanometer across
both graphics, client, and our semi-custom business."
"we are 28 this year"

I understand that English isn't your first language, but still, that's pretty cut-and-dry. Perhaps your difficulty with English is part of what's causing you issues? You literally proved yourself completely wrong.

And, again, you're trying to look for things which don't deny 20nm this year instead of things which say it's coming this year. You definitely have a twisted understanding of what facts and sources are.

 

[Silverforce11]

Its pretty damn clear they have 20nm IN DESIGN (marketing for: therefore not anywhere close to launch) and finfets thereafter (much later). If they are launching this year they will hype it up much more.

 

[NostaSeronx]

AMD released this in January 2013;

Huh, what is that in the distance. There is no new GPUs in 2013, oh snap!

March 2013;
http://www.pcper.com/reviews/Graphi...eview-Sea-Islands-and-Bonaire-Make-Appearance

September 2013;
http://www.pcworld.com/article/2049397/amd-unveils-hawaii-generation-of-gpus.html

But, wait the roadmap doesn't show Q4 2013. Well, both GPUs launched and leaked way before that. Then, you have the codenames being announced in Q3 2013. So much for stable throughout 2013.

You also have Oland replacing the VLIW4 architectures in 2013.

 

[Techhog]

AMD released this in January 2013;

Huh, what is that in the distance. There is no new GPUs in 2013, oh snap!

March 2013;
http://www.pcper.com/reviews/Graphi...eview-Sea-Islands-and-Bonaire-Make-Appearance

September 2013;
http://www.pcworld.com/article/2049397/amd-unveils-hawaii-generation-of-gpus.html

But, wait the roadmap doesn't show Q4 2013. Well, both GPUs launched and leaked way before that. Then, you have the codenames being announced in Q3 2013. So much for stable throughout 2013.

You also have Oland replacing the VLIW4 architectures in 2013.

And? You're still treating a possibility of something happening as a confirmation that it will. Unless you're trying to say that AMD always lies, you're not proving any facts here.

 

[bunnyfubbles]

What AMD needs to do is fix their cut outs and port placement on the back of their blowers. Move the Dvi ports like Nvidia and switch to this:

If only Dvi could get phased out, such massive connectors for what it can do.

DVI will be phased out very soon its pretty much incompatible for next gen monitors. 4K@60Hz needs more than dual link DVI, and ASUS' new 1440p144Hz G-Sync monitor only has Displayport as an option.

DVI-D still viable for 1440p60 and 1080p120, but I have two such monitors that are already a couple years old each (Dell U2711 and BenQ XL2420T) and each already have Displayport as an option despite being a couple generations old...

We're basically at that point where if you don't have one of these relatively modern monitors that have DP connectivity, these next gen high end GPUs simply aren't for you anyway (although I wouldn't count on them ditching DVI altogether, but we're definitely at that point where they can stop including more than 1)

 

[NostaSeronx]

Unless you're trying to say that AMD always lies, you're not proving any facts here.

AMD will not ever officially talk about a device until launch. Even if there was substantial evidence that these devices will launch. Until launch AMD will spin(to win) and give answers that are not definitive. Everything must seem like a surprise to incite monetization.

HBM - Ready since Q3 2014. (July)
2.5D TSV - Ready since Q2 2014. (April - June)

The only puzzle piece left is 20-nm LPM. Which we should be seeing volume production announcements for anytime soon.

 

[AtenRa]

AMD's last quarter transcript conference call on seeking alpha. Look it up.

heh, i just realized he was talking about this year.

 

[Techhog]

AMD will not ever officially talk about a device until launch. Even if there was substantial evidence that these devices will launch. Until launch AMD will spin(to win) and give answers that are not definitive. Everything must seem like a surprise to incite monetization.

HBM - Ready since Q3 2014. (July)
2.5D TSV - Ready since Q2 2014. (April - June)

The only puzzle piece left is 20-nm LPM.

There's a huge part of your puzzle missing here: rumors. Every rumor has either been fake or pointed to 20nm in 2014. With Bonaire and Hawaii, we had leaks to confirm them. We have no leaks or rumors pointing to 20nm this year. When a rumor appears pointing to 28nm only this year, you say that it's wrong, calling your speculation fact. There's no way to argue with you because you've decided that you have to be right. If something proves you wrong, you call it wrong because you know that you're right. You're either the most stubborn person on the planet, or you don't know what a fact is. You have no facts other than "AMD is working on 20nm GPUs". All of us (except Shintai I guess) know this. What we don't know is exactly when, but it's most likely next year.

You're saying that it's a fact that it'll launch this year because AMD denied it. What does that mean? If they had confirmed it, does that mean that it wouldn't be on the way? Of course not! So, basically, you are saying that, no matter what AMD says, you're right. You have no basis. You just believe this to be true, which means that it's true. Your mindset is completely twisted. Face it, you don't know how facts work. Go grab a dictionary. If you need to put things together and make guesses to figure something out, it's not a fact; it's speculation.

 

[stahlhart]

Enough of this.
-- stahlhart