Fudzilla: New AMD Zen APU boasts up to 16 cores (plus Greenland GPU with HBM)

[R0H1T]

I think its you being the one with the problem.

None of the 2 you link to in the post claims, predict or state anything ridiculous.

The answer to both your statements is the same. You (I mean the OP) know that BDW -> Skylake will not be the same as P4 -> Core & that's (I'd say the whole thread) as close to predicating, such hypothesis yourself, as it is to trolling here on AT. The same goes for 65W BDW beating the Devil's Canyon, now you could either define them as statements, predictions, guesstimate (or hope) perhaps even trolling, as some would. Pick your poison but the argument that one camp overhypes their products while the other doesn't is a fallacy & the same goes for their supporters.

 

[Enigmoid]

They didn't really hide this from anyone. Actually Lisa wrote that the future is about focused ARM offers "ARM only where it makes sense", far different from the "ARM will win in the long run" statement that came under the duo Feldman/Read. She was also pretty blunt when talking about Skybridge, telling everyone that it didn't generate enough interest on the market, and she did not commit to Seattle ramp up despite the product being ready. This is as direct as she could be, especially because most of the team that was executing the old plan is still at AMD, and the BoD is the same.

Why not go and talk to the market and OEMs and see if something like Skybridge is a good idea first before putting it on the roadmaps and devoting money and time towards it? Seems like a management error.

There are. One is for the same reason Intel couldn't keep Conroe clocks at Prescott levels: A beefier core is harder to clock higher than an anemic core like Bulldozer, and the other is that the 14nm Samsung node isn't really optimized for this kind of clocks.

Conroe overclocked just as high.

http://www.anandtech.com/show/2045/18

~3.5 ghz with some chips reaching 4.0 ghz

Zen could overclock well and reach high clocks but I doubt AMD will try to go all out for > 4 ghz clocks.

 

[ShintaiDK]

The answer to both your statements is the same. You (I mean the OP) know that BDW -> Skylake will not be the same as P4 -> Core & that's (I'd say the whole thread) as close to predicating, such hypothesis yourself, as it is to trolling here on AT. The same goes for 65W BDW beating the Devil's Canyon, now you could either define them as statements, predictions, guesstimate (or hope) even trolling, as some would. Pick your poison but the argument that one camp overhypes their products while the other doesn't is a fallacy & the same goes for their supporters.

You are getting desperate. Nobody claimed anything.

This is from your post:

I do hope that they reveal/leak the specs soon, I'm eager to know if it will be a faster chip than the 4790k, or if we'll have to wait for Skylake next year.

 

[R0H1T]

You are getting desperate. Nobody claimed anything.

This is from your post:

And you know the answer to the post, probably the poster know it too so what would you call that?

 

[mrmt]

Why not go and talk to the market and OEMs and see if something like Skybridge is a good idea first before putting it on the roadmaps and devoting money and time towards it? Seems like a management error.

There was interest in Skybridge-like SKUs circa 2012 when AMD developed this strategy, but since then things changed a lot on the Intel front, with HSW-EP servers bringing very efficient SKUs, Atom-C and now Xeon-D. The buzz around ARM servers faded around the beginning of 2014. This is why Nvidia, Samsung and others backtracked from their strategy and the small players either shrank or went out of business. The notable exception is Qualcomm, but they have the luxury of time and money to slowly probe this market to see whether this is a viable one.

AMD was a notable exception here, too small to adopt Qualcomm's wait-and-see approach and with a management team too invested in this strategy to back off, so they pressed further until Rory's dismissal.

That they subordinated their consumer strategy (which pays their current bills now) to their server strategy (a mirage for them after K8 days) is a testament on how deeply invested Rory's team was on this vision, and I'm afraid that the company will pay a heavy price once Zen arrives on the market being unsuitable for the consumer market.

Conroe overclocked just as high.

http://www.anandtech.com/show/2045/18

~3.5 ghz with some chips reaching 4.0 ghz

Zen could overclock well and reach high clocks but I doubt AMD will try to go all out for > 4 ghz clocks.

The overclocked chip, and this after Intel tweaked the process. Stock chips, especially the initial batch, featured much lower clocks. Unless we are talking about 300W water cooled SKUs, I don't think Zen will get even close to 4GHz

 

[Shehriazad]

Does AMD even NEED to go /above 4 Ghz base with Zen?


Please correct me if I'm terribly wrong, but here goes nothing:

Haswell-E's fastest base clocked at 3.5 Ghz. (or did I miss a new release?)
Excavator + 40% would be like what...10% within the IPC margin, no?

3500 + 10% = 3850. So AMD only needs to be able to reach 3850 to be dead even...no?
Ivy Bridge-E would be BEAT if AMD managed 3900 base and 4100 boost, eh?


Of course I'm NOT saying that this is a fact...nor that this WILL happen...I'm just saying...4 Ghz base is totally not a requirement with an IPC gain this massive...but keep in mind that AMD is all about the high clocks.

Also keep in mind that AMD doesn't really mind their TDPs...I expect their 14nm parts to still be at 95W.
So there is plenty of headroom to deliver their typical 3.5-4 Ghz chips.


So basically...if it's Excavator + 40% @ 3.9 to 4.0...Haswell-E and Ivy-E are "beat" and Intel will actually have to compete with their next CPU..that will probably have anywhere from 10-15% more performance...thus again AMD will probably be a tiny step behind. Buuuuuuuut...they could also surprise us and release their quads @ 4.1 Base with 4.3 boost or something like it. (Their FX4350 had a 4.2 base...different arch of course...but AMD tries to hammer dem clocks. I'm leaving out the FX9000 series because that series is not "your normal CPU")

 

[frozentundra123456]

They didn't really hide this from anyone. Actually Lisa wrote that the future is about focused ARM offers "ARM only where it makes sense", far different from the "ARM will win in the long run" statement that came under the duo Feldman/Read. She was also pretty blunt when talking about Skybridge, telling everyone that it didn't generate enough interest on the market, and she did not commit to Seattle ramp up despite the product being ready. This is as direct as she could be, especially because most of the team that was executing the old plan is still at AMD, and the BoD is the same.

I think the show was all about dismantling "ambidextrous" and going back to x86, but it also showed a weakness: K12 was supposed to be the lead product and since it was server optimized it also made Zen being server optimized. Basically AMD will again bet their farm on a server optimized architecture on the consumer market while lacking the funds optimize for it.

For AMD current strategy they would be far better suited to have a consumer optimized architecture and server products to get the low end of the server market, think a Xeon-D competitor + APU line leveraging on AMD graphics. But it seems that "ambidextrous" damaged this strategy and now AMD has to go after medium servers and *desktops*.

Yep, forcing a server optimized architecture down into the consumer market. Seems eerily familiar doesnt it, even down to the outrageous predictions with no solid information. Maybe they will do it right this time. At least they cant lose the server market this time around because they have already done that.

 

[piesquared]

He's talking about the Zen based FX series.

It has no iGPU.

With a 40% increase in IPC, ZenFX would totally disrupt the PC gaming market. In fact it's what enthusiasts have been asking for forever. A dedicated enthusiast gaming CPU driving a blazing fast dedicated GPU in the low level API era where all available cores can be utilized efficiently has all the ingredients to be a best seller. APUs slot in below that, just like they have always done.

 

[mrmt]

Yep, forcing a server optimized architecture down into the consumer market. Seems eerily familiar doesnt it, even down to the outrageous predictions with no solid information. Maybe they will do it right this time. At least they cant lose the server market this time around because they have already done that.

The problem this time is that they are already saying that they *will* lose some of the consumer market. Now it's only a question of how much they are going to lose.

 

[mrmt]

With a 40% increase in IPC, ZenFX would totally disrupt the PC gaming market. In fact it's what enthusiasts have been asking for forever. A dedicated enthusiast gaming CPU driving a blazing fast dedicated GPU in the low level API era where all available cores can be utilized efficiently has all the ingredients to be a best seller. APUs slot in below that, just like they have always done.

Sorry, enthusiasts have been asking for cheaper Xeon 8C SKUs, not more AMD hot air. We already have plenty of that in the last 7 years.

 

[R0H1T]

The problem this time is that they are already saying that they *will* lose some of the consumer market. Now it's only a question of how much they are going to lose.

That's most the likely the 4 core APU desktop segment & below, also the laptops. If anything the high end desktops could be their only saving grace next year, at least till they release something with HBM on a desktop or most likely a notebook.

 

[mrmt]

That's most the likely the 4 core APU desktop segment & below, also the laptops. If anything the high end desktops could be their only saving grace next year, at least till they release something with HBM on a desktop or most likely a notebook.

My prediction is that they will lose a lot in the notebook market, won't be able to establish themselves on the tablet market, will get a small share on the web-hosting server market and will become a cheap second source to Intel on the desktop market with a significant presence there, with a caveat. If what you understand as a high end desktop is workstation-level Xeon SKUs, they won't get there. If you think Intel mainstream CPUs, then yes, they have a shot.

 

[piesquared]

Sorry, enthusiasts have been asking for cheaper Xeon 8C SKUs, not more AMD hot air. We already have plenty of that in the last 7 years.

Looks like intel blew it then. They'll be able to get the ZenFX instead.

 

[Erenhardt]

My prediction is that they will lose a lot in the notebook market, won't be able to establish themselves on the tablet market, will get a small share on the web-hosting server market and will become a cheap second source to Intel on the desktop market with a significant presence there, with a caveat. If what you understand as a high end desktop is workstation-level Xeon SKUs, they won't get there. If you think Intel mainstream CPUs, then yes, they have a shot.

Lottery numbers plz

 

[R0H1T]

My prediction is that they will lose a lot in the notebook market, won't be able to establish themselves on the tablet market, will get a small share on the web-hosting server market and will become a cheap second source to Intel on the desktop market with a significant presence there, with a caveat. If what you understand as a high end desktop is workstation-level Xeon SKUs, they won't get there. If you think Intel mainstream CPUs, then yes, they have a shot.

Yes they will, certainly in the short to medium term but for long term the eventual transition to Zen/K12 or whatever else they'll put in the notebooks is going to decide the future i.e. if AMD will be relevant in that part of the market or not. I will however add that HBM could be a major game changer for AMD in this segment as they'll always retain a lead over Intel in the GPU's but can dedicate some more die space/power for their CPU, exact opposite of what Intel is doing.

High end desktops can & probably will accept Zen readily if it's a nice performer, remember all the perf/watt restrictions don't apply in this segment & so long as they don't make the 8 core chip as hot as the 9590 (TDP of 220 W) it should sell. Also they ideally they need a 4 core Zen, sans IGP, to crack the 200~400$ mainstream desktop segment as the current APU's simply won't do.

 

[Phynaz]

The problem this time is that they are already saying that they *will* lose some of the consumer market. Now it's only a question of how much they are going to lose.

Consumer market is shrinking. I don't see where they are stating they will lose share.

 

[frozentundra123456]

Or maybe those predicting 8 core zen at 95 watts at 4ghz with 40% ipc increase for 200.00 could send *me* some lottery numbers. Or maybe the ones who were predicting that 16 core zen apu with HBM in the thread title could send me some. Oh wait, guess those might be wrong on all counts.

 

[mrmt]

Looks like intel blew it then. They'll be able to get the ZenFX instead.

We could also get Bulldozer FX, that doesn't say anything about the quality of it, does it?

In any case, the next several quarters before Zen introduction will be pretty interesting. Kumar got a lot of heat because of his stability forecast on the consumer market until Zen Introduction and by AMD's own admission they are also forecasting more personnel cuts (albeit a small one) for the second half of the year. I wonder what happens if Carrizo sales mimic the behavior of its previous CMT ancestors and keep crashing, we might even have further cuts. With AMD there's always a chance of a product delay, as usually happens with this company. That will put them in even greater strain.

Interesting times ahead, and plenty of opportunities to make money too.

 

[cytg111]

Or maybe those predicting 8 core zen at 95 watts at 4ghz with 40% ipc increase for 200.00 could send *me* some lottery numbers. Or maybe the ones who were predicting that 16 core zen apu with HBM in the thread title could send me some. Oh wait, guess those might be wrong on all counts.

Instead of semi-quoting "those" and "predicting 8 core zen at 95 watts at 4ghz with 40% ipc increase for 200.00" you could find some actual references to these facts? I remember alot of maybes and speculation - I dont remember so many precise statements or predictions. We do like our facts straight, right?

 

[Snafuh]

Yes its not directly marketshare. But its revenue and margins. And obviously AMD is not going to sell cheap chips for pennies to gain any share due to margins.

PC shipments are declining. AMD can keep the revenue flat and have a little market share gain because the market is shrinking.
First you said Zen might be cancelled, then you predicted Zen will be a low performance core. Both was wrong yet you can't acknowledge that AMD might have a good product in development.

 

[Abwx]

Lottery numbers plz

Toss a coin, you ll get much more accurate predictions than from any of thoses experts in everything that flood some threads..

High end desktops can & probably will accept Zen readily if it's a nice performer, remember all the perf/watt restrictions don't apply in this segment

Mark Papermaster stated that they wont backtrack from efficency with Zen despite a "doubling of the core performance caracteristic on Zen"...

 

[mrmt]

Consumer market is shrinking. I don't see where they are stating they will lose share.

They said they will lose sales. And if the best case scenario is a slightly growth in sales, coupled with the low starting number, that means AMD isn't having much hopes for its products on the consumer market.

 

[R0H1T]

Mark Papermaster stated that they wont backtrack from efficency with Zen despite a "doubling of the core performance caracteristic on Zen"...

Yes I understand that but we don't know if the 8 core Zen will be more efficient than an 8 core Intel, sure it'll be more efficient than AMD's very own BD but that wasn't my point

 

[AtenRa]

They said they will lose sales. And if the best case scenario is a slightly growth in sales, coupled with the low starting number, that means AMD isn't having much hopes for its products on the consumer market.

They said they are not after high volume low profit segments/products, nobody said they will lose sales but you.

 

[Qwertilot]

If you stop chasing high volume/low margin products its surely basically axiomatic that you lose sales? Might gain profit though