Samsung Will Begin Manufacturing AMD's 14nm Zen Core By Year's End

[patrickjp93]

They play both sides for the console war.  The console wars should make them lots of monies.  Hopefully,  AMD should be very competitive against Intel and, Nvidia for the next couple gens.  

The console wars? Pal you can fit 52x the number of chips made for consoles into the number made for PCs, not to mention phones, embedded systems, routers, cars, servers, supercomputers, etc. It's publicity for AMD, but trust me, it wasn't as much money as you think.

[dalekphalm]

So he made a few dual-cores...woopdy do... He's been working at Apple for more than 7 years now, and all Apple makes are ARM chips for phones, effectively embedded SOCs. Do you really need a source for common knowledge?

No, you misunderstand. I was not asking for a source for him working at Apple for 7 years. We all know this. I was asking for a source because you kept referring to him as "Just a SoC designer". You were completely ignoring (Or perhaps you didn't know?) his previous history with AMD, x86, and AMD64. Sure this doesn't magically make him the God King of CPU design, but you implied he had only worked with ARM design. His history is quite well packed with success.

 

Of course only time will tell whether this will be awesome or another flop. But I'm willing to give them the benefit of the doubt based on Keller's track record. I will wait for reviews before I condemn AMD.

[patrickjp93]

No, you misunderstand. I was not asking for a source for him working at Apple for 7 years. We all know this. I was asking for a source because you kept referring to him as "Just a SoC designer". You were completely ignoring (Or perhaps you didn't know?) his previous history with AMD, x86, and AMD64. Sure this doesn't magically make him the God King of CPU design, but you implied he had only worked with ARM design. His history is quite well packed with success.

 

Of course only time will tell whether this will be awesome or another flop. But I'm willing to give them the benefit of the doubt based on Keller's track record. I will wait for reviews before I condemn AMD.

It won't be a flop. AMD is finally getting rid of the travesty that was CMT, going instead with SMT, and Keller's forte' is in IPC optimization, though it's been a long time since he worked on the rich instruction set of x86.

 

And as I said above, he made dual-cores when it was easy to make dual-cores, in less than 300 million transistors. Since then he's spent every waking day at Apple working on ARM SOCs. I'd expect his skills at dedicated CPU design to be a little rusty.

[Humbug]

Ever hear the phrase "If you don't use it, you lose it?" Back when you had less than 300 million transistors he had success. Back when it was really damn easy, he had success.

lol. You think designing an Athlon 64 era CPU is damn easy?

[Jauris]

Just made the switch to Intel after a number of years with AMD, but I'll gladly switch back over in a heart beat if they can compete. I'll definitely be keeping my eye on this.

[Linux User]

Hopefully this will give an option to broke enthusiasts like me. Either dish out heaps of money for Intel or go for sub par single threaded performance with AMD. This is FXing great news.

[Linux User]

AMD is back on the game

[patrickjp93]

lol. You think designing an Athlon 64 era CPU is damn easy?

Relatively yes. The laws of quantum physics weren't a problem in the days of 50 some nm nodes. And, considering Intel whooped their butts less than 2 years later, do you really think Intel is that vulnerable?

[Builder]

Relatively yes. The laws of quantum physics weren't a problem in the days of 50 some nm nodes. And, considering Intel whooped their butts less than 2 years later, do you really think Intel is that vulnerable?

Pardon me for being an idiot but doesn't the entire function of silicon semiconductors rely on quantum physics?

[patrickjp93]

Pardon me for being an idiot but doesn't the entire function of silicon semiconductors rely on quantum physics?

Above 50nm most of it is still governed by macro physics or classical E&M. This is because the Shrodinger wave function and Heisenberg's uncertainty have no bearing on systems above a certain scale.

[That Norwegian Guy]

I think AMD has to target the niche gap in between the consumer i7 and the lower end extreme editions. 6-8 (proper) cores and a bigger die size to make up for the node disadvantage. Then be as competitive as possible with the i7 pricing.

 

It will be a more expensive AMD chip than we've probably ever seen, but it'll roflstomp Intel's conservative "choose between 4 core or 4 core with HT" dichotomy and thus people will buy it. If it's more like the piledrivers or bulldozers they can just grind them down to dust already for all I care. As it stands, there's a freaking ATOM (!) chip capable of competing with an 8350.

[ComradeHX]

Hopefully something to compete with Intel.

 

However, knowing AMD; it probably won't.

It will compete with intel.

But as usual(for these few years), the only place AMD wins is price and function as room heater.

[mr moose]

As much as this sounds good, I will not count any chickens before they hatch.  

[mr moose]

Hopefully something to compete with Intel.

 

However, knowing AMD; it probably won't.

 

I really like the way everyone seemed to jump on this comment like he doesn't know anything, for anyone who might have taken he time to look at his profile he was 7 years old when Jim Keller  worked on the Intel killer CPUs.  So it's pretty fair to assume his comments come from his experience of AMD products over the last 4 years or so, and on that note they actually make sense. 

 

Well done LTT for assuming he's being a troll and reacting like fanboys.

[pixeldensity]

I really like the way everyone seemed to jump on this comment like he doesn't know anything, for anyone who might have taken he time to look at his profile he was 7 years old when Jim Keller  worked on the Intel killer CPUs.  So it's pretty fair to assume his comments come from his experience of AMD products over the last 4 years or so, and on that note they actually make sense. 

 

Well done LTT for assuming he's being a troll and reacting like fanboys.

well looking at the past post only three out of the whole entire thread was poking fun of him, so its not the majority of the whole LTT being fanboys. If you want to see a real fanboy look two post above your last post. Plus you can't expect everyone to look through profile to see when he was born.

 

Well getting back on topic

 

 

As much as this sounds good, I will not count any chickens before they hatch.  

I do agree with you, we should wait till the product releases first to see if it anything worth celebrating. So we shouldn't over hype zen. But with the current news that coming out about zen and with Jim Keller at the helm, thing do look promising.

[monstercameron]

well played samsung neutralize intel forum troll fanboys http://www.eetasia.com/ART_8800704907_480200_NT_b00975ae.HTM

[Paragon_X]

Interesting, lets see what this will bring us.

 

Also does anyone else thinks that Skybridge is kind of pun for Skylake,

[Humbug]

It will be a more expensive AMD chip than we've probably ever seen, but it'll roflstomp Intel's conservative "choose between 4 core or 4 core with HT" dichotomy and thus people will buy it. If it's more like the piledrivers or bulldozers they can just grind them down to dust already for all I care. As it stands, there's a freaking ATOM (!) chip capable of competing with an 8350.

I'm guessing another 1st gen bulldozer would bankrupt AMD. They can't afford to do that again.

[mr moose]

well looking at the past post only three out of the whole entire thread was poking fun of him, so its not the majority of the whole LTT being fanboys. If you want to see a real fanboy look two post above your last post. Plus you can't expect everyone to look through profile to see when he was born.

 

Well getting back on topic

 

 

 

 

be that as it may, I do expect people to think before they post, Even without looking at his profile it's easy to see he was referring to recent history and even then he wasn't being derogatory toward AMD. 

 

And I am also getting really tired of fanboys shooting down anyone who posts an honest personal opinion as if they are trying to deceive the world.  It has been happening a little too often of late.

[LAwLz]

Don't get your hopes up.

Keep your expectations as low as possible so you won't get disappointed when they don't live up to the ridiculously high expectations people are already forming.

It's better to have your low expectations exceeded than to have your high expectations not met.

You don't want to end up with another Bulldozer...

[Eniqmatic]

im talking CPU performance. if the boards were shit, thats a different question... i still know a ton of people running phenom2s OCd no problem. youre just picking on the bad things of AMDs and not looking at what intel did crappy then (granted i cant remember atm, since its too late, but i know they had their fair share of problems). when threads like these arise, you are such an intel fanboy...

 

Still have a machine with a 1100T in it, loved it back when I first got it. 

[vm'N]

Jim Keller is an embedded SOC designer. Don't get ahead of yourself.

Jim Keller have also designed high performance CPUs before.

 

And AMD beat Intel to exactly 2 things: 64-bit

AMD made x86-64. Intel already had a functional 64bit processor working. IA-64

 

It's obvious already that CPU monopoly is all about IPC and AMD cpu's we've seen last 10 years don't even have more IPC than Core 2.

What? IPC in itself explain nothing about the performance of a processor. IPC is a stupid messurement.

I wonder how IPC meassure the difference between integer and float calculations, between predictable and non-predictable code, between loops and whatnot.

 

They don't. The only other company which builds large chips is IBM. Keller is a bit of a wild card, but he's in the same boat having only ever designed embedded SOCs.

No.

 

So he made a few dual-cores...woopdy do... He's been working at Apple for more than 7 years now, and all Apple makes are ARM chips for phones, effectively embedded SOCs. Do you really need a source for common knowledge?

It is not the point if him making something that would look obsolete today, but the fact that he brought new ideas to the industry.

Apple processors have also become incredible advanced, especially their cyclone architecture.

 

 

It won't be a flop. AMD is finally getting rid of the travesty that was CMT, going instead with SMT, and Keller's forte' is in IPC optimization, though it's been a long time since he worked on the rich instruction set of x86.

 

And as I said above, he made dual-cores when it was easy to make dual-cores, in less than 300 million transistors. Since then he's spent every waking day at Apple working on ARM SOCs. I'd expect his skills at dedicated CPU design to be a little rusty.

CMT in itself it not bad. It was AMDs implementation of it that was bad. They simply didn't have the technology to do what they wanted.

"Rich instruction set of x86", people normally would call it the bloated instruction set (depend on how you look at it).

 

I think AMD has to target the niche gap in between the consumer i7 and the lower end extreme editions. 6-8 (proper) cores and a bigger die size to make up for the node disadvantage. Then be as competitive as possible with the i7 pricing.

That would be a ridiculous target. K12 and Zen will open up for a lot of markets for AMD. Remember that AMD will have multiple products running the same architecture. Just like Intel have multiple productlines running the same architecture. They cannot only produce 1 architecture for a single productline.

Core-count is still unknown.

 

It will compete with intel.

But as usual(for these few years), the only place AMD wins is price and function as room heater.

Are you talking about AMDs current or upcomming architecture? Where do you known anything about AMDs upcomming pricings and temperatures?

[patrickjp93]

@vm'N The days of K7&8 are not exactly high-performance days. Also, We consider Itanium functional?

IPC is not a stupid measure, especially when you have the cycle count for instructions confirmed by the manufacturer as in Intel providing the timing on floating point multiply/divide. Considering we know both AMD and Intel support near-equivalent instruction sets, we can use IPC perfectly well.

[vm'N]

@vm'N The days of K7&8 are not exactly high-performance days.

For that time it were.

Also, We consider Itanium functional?

Itanium is functional. Hence why it is still around. It however is simply not the replacement for x86 as Intel orginally thought.

IPC is not a stupid measure, especially when you have the cycle count for instructions confirmed by the manufacturer as in Intel providing the timing on floating point multiply/divide. Considering we know both AMD and Intel support near-equivalent instruction sets, we can use IPC perfectly well.

IPC is an incredible stupid measurement for performance. It however provide an idea of the effeciency of an architecture, but do not describe the performance of an architecture.

Because what about frequency, code optimization, dataraces, misprediction, cache miss, and much much more.

Also x86 takes "macro" instructions, also known as a complex instruction. This complex instruction will then be translated into AMDs/Intels own RISC-a-like code.

A single complex instruction can be translated into multipe RISC-a-like instructions.

[DigitalHermit]

*grabs chips/crisps*

 

I'm just sitting here waiting for the chips to hit the entry-level market and hold judgment when they do...