• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Zen ETA Q4, '16 ?

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Fjodor2001 said:
See answer by AtenRa. Also, when was the last time Intel released a new CPU on schedule? If anything they have been plagued by delays the last few years, and it continues with Cannonlake and 10 nm.
Click to expand...

I didn't know AMD schedule had anything to do with Intel and Cannonlake, especially because 10nm is still a wishful thinking for them. But still, when did AMD launch a processor on time?
 
AMD said that only server parts will be available in 2016 and that the mainstream APUs won't be available until 2017. Makes sense if they are assuming there will be yield issues.
 
Idontcare said:
They aren't, which can be a huge problem (for GF and their customers).

Cut and paste would require "copy exact", which is more costly in terms of equipment costs than what GF is attempting.

GF wants to recycle as much of their existing equipment from prior nodes as possible, purely for cost savings (penny wise, pound foolish) which has resulted in them cobbling together something that is similar to Samsung's process ("electrically equivalent" is the vernacular we use) but the pain comes in terms of process variability (parametric yield loss) and defectivity (functional yield loss).

On the flipside though, it is GF's only hope for a future, so eventually they have to get it right or they risk becoming the next Jazz or Tower.
Click to expand...

Wow, that is a total lack of foresight on AMD and GF for allowing this to happen. Intel generated 13 billion in sales last quarter. If AMD has a competitive product that can recapture even a small portion of that market share, they are losing billions for every quarter they delay.
 
Arachnotronic said:
Any insight into whether the DigiTimes report re: GloFo's 14nm yields is true? Thanks!
Click to expand...

No insight as to whether or not GF's 14nm yield is the specific cause of a delay on AMD's part to get Zen off the ground...yes their yields are craptastic but the timeline involved here is nearly 1.5 years out, which makes me suspect the yield argument is a red herring.

What I would be more apt to believe is that AMD has so woefully underfunded the R&D for Zen such that the tapeout date has now been pushed out far enough so as to make a Q4'16 ship date virtually impossible to hit (regardless of yields).

But it is a lot easier to blame the foundry because the foundry can't come back and make a rebuttal to the contrary, bad for business to show-up the customer.
 
positivedoppler said:
Wow, that is a total lack of foresight on AMD and GF for allowing this to happen. Intel generated 13 billion in sales last quarter. If AMD has a competitive product that can recapture even a small portion of that market share, they are losing billions for every quarter they delay.
Click to expand...

It costs money. AMD doesn't want a customized process node, they want a bog-standard node because they can't pay for the extra R&D.

And think about GF, there is a reason they don't have their own 14nm process node, that reason is money. They can't afford to field their own process node R&D team, so they had to license Samsung's process node.

Since they found themselves in this situation because of choices they had to make based on their economics, it isn't too surprising to see that be a continuing theme in how they decided to bring Samsung's 14nm in-house. Cheap is as cheap does, either out of desperation or because they just aren't in the position to shoot for doing any better.
 
Phynaz said:
Is that a new CPU, or a respin of an existing CPU?
Click to expand...

It's a respin with a new stepping. But you asked about when they met their launch date, and they met it, at least with the 7870k anyway.

The 870k, despite having been announced already, is not available for purchase. They may have the 880k out by the time they actually start selling the 870k.
 
As far as i know, GloFo is already producing 14nm LPE wafers in Fab 8 at New York.

What AMD will use for ZEN and any other 14nm products is the 14nm LPP in Fab 1 in Dresden Germany. The 14nm LPP was originally scheduled for late 2015 early 2016 production.
But, ZEN is not the only product for 2016 and we may see 14nm products before ZEN release.
 
AtenRa said:
As far as i know, GloFo is already producing 14nm LPE wafers in Fab 8 at New York.

What AMD will use for ZEN and any other 14nm products is the 14nm LPP in Fab 1 in Dresden Germany. The 14nm LPP was originally scheduled for late 2015 early 2016 production.
But, ZEN is not the only product for 2016 and we may see 14nm products before ZEN release.
Click to expand...

http://globalfoundries.com/technology-solutions/leading-edge-technology/14-lpe-lpp

14nm LPE/LPP is at GF's Fab 8 in NY Malta. :thumbsup:

14nm LPP will also be the most advanced GF process with support for 2.5D/3D stacking (crucial for AMD's 2017 server APU)

https://www.semiwiki.com/forum/content/4567-globalfoundries-14-15-8-a.html
 
Last edited:
DrMrLordX said:
It's a respin with a new stepping. But you asked about when they met their launch date, and they met it, at least with the 7870k anyway.

The 870k, despite having been announced already, is not available for purchase. They may have the 880k out by the time they actually start selling the 870k.
Click to expand...

No, I asked when they met their date with a new CPU, not a stepping and a speed bump.
 
Phynaz said:
Good
Click to expand...

So will Bristol Ridge count as a new CPU? It's 4 Excavator cores with 512 GCN 1.2 shaders on AM4, and at this rate, AMD might have to launch it and AM4 early to cope with Zen delays. But, you know, we can't give them any credit for launching it on time (or early), since it's just Carrizo in a new package and (probably) without HDL libraries. Right?

Or am I missing something here?
 
DrMrLordX said:
So will Bristol Ridge count as a new CPU? It's 4 Excavator cores with 512 GCN 1.2 shaders on AM4, and at this rate, AMD might have to launch it and AM4 early to cope with Zen delays. But, you know, we can't give them any credit for launching it on time (or early), since it's just Carrizo in a new package and (probably) without HDL libraries. Right?

Or am I missing something here?
Click to expand...
Supposedly Bristol Ridge could enter on 14 nm instead on 28 nm.. But that seems to be discarded... So AMD will hardly have some performance increase.
 
dark zero said:
Supposedly Bristol Ridge could enter on 14 nm instead on 28 nm.. But that seems to be discarded... So AMD will hardly have some performance increase.
Click to expand...

die shrink with same uarch i think is a new CPU. especially with regular rather than HD libraries.

sounds like AMD's kabylake.
 
Fjodor2001 said:
Since we're discussing what Intel CPUs that AMD Zen will compete with, I think it does.
Click to expand...

But Sandy Bridge and Sandy Bridge-E were released 4 years ago, so the timing of their release dates vs. projections is somewhat irrelevant at this point.
 
Idontcare said:
No insight as to whether or not GF's 14nm yield is the specific cause of a delay on AMD's part to get Zen off the ground...yes their yields are craptastic but the timeline involved here is nearly 1.5 years out, which makes me suspect the yield argument is a red herring.

What I would be more apt to believe is that AMD has so woefully underfunded the R&D for Zen such that the tapeout date has now been pushed out far enough so as to make a Q4'16 ship date virtually impossible to hit (regardless of yields).

But it is a lot easier to blame the foundry because the foundry can't come back and make a rebuttal to the contrary, bad for business to show-up the customer.
Click to expand...

If you are using duct tape, clamps and cardboard to align your equipment 1.5 years is not a long time ;-) There's a reason Intel spends so much on R&D and equipment doing it right costs a crap ton of money.
 
svenge said:
But Sandy Bridge and Sandy Bridge-E were released 4 years ago, so the timing of their release dates vs. projections is somewhat irrelevant at this point.
Click to expand...

Who has mentioned SB and SB-E?

I was talking about the Intel releases the last few years and the delays we've seen for those, and what we can expect going forwards (CannonLake and 10 nm delayed, KabyLake stopgap solution introduced).
 
Last edited:
dark zero said:
14 nm and using Carrizo improvements... Seems Phenom II all over again.
Click to expand...

Not sure what you mean, but Phenom II was a pretty reasonable low-cost alternative to the Intel processors at the time.

Their cut-down siblings, Athlon II X4, were very good deals. Built a ton of systems for people around them.
 
Some people here seem to have too much spare time for repeating the same arguments over and over about well known facts. To follow this tradition let me quote myself again:
A few weeks ago I found an AMD'ers LinkedIn profile mentioning, that he works on the next gen x86 arch to arrive in 2017.

Edit:
The exact wording:
System Architect for Carrizo APU (launched in 2015) and AMD's next-generation (to be launched in 2017) mainstream mobile/desktop FX and A-series APU.
Click to expand...
Click to expand...
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top