AMD’s Next Gen x86 High Performance Core is Code Named “Zen”

http://wccftech.com/breaking-amds-gen-x86-high-performance-core-code-named-zen-debut-k12/

CHOO CHOO!!! HERE COMES THE HYPE TRAIN!!!

 

gonna flop - you heard it here first

 

we,ll get back to you in 2016 when they release it
but honestly with that dude running it I can see it doing well(well at least have a chance)

problem is in the time it takes them to get it on the shelves what will inhell come up with.

 

Intel hasn't really pushed the envlope recently it seems, 5% here and there.

Bring back the old AMD, underdog for life!

 

They need to drop the 'A' from AMD.

False advertising is what it is!

 

Hasn't really been much need. The cheapest LGA1155 Celeron is way, way more than enough processing power for most people these days. Even for gaming, a mid-range Core i5 does a perfectly good job. Better to save on the CPU and get a quick GPU and SSD instead.

If AMD can beat Intel's top-end CPUs then that'll be great marketing, but (as always) what they actually need is something that competes with Intel's mid-range. Better performance with lower cost and lower power is critical.

Of course, if they can repeat the K8 and offer a mid-range CPU that out-performs Intel's top-end, with lower-power than any non-mobile Intel CPU, that'll be even better.

 

They dug up Steve Jobs and started a campaign of making amd look cool, apple tactics?

 

2016!! to release a CPU that has 25% more performance then that they have on sale now! IDIOTS!! They are all idiots before they waste another 2 years on a FLOP!!

AMD should “just die shrink of Phenom II to 20nm, make it 8 core and sell it”

People have been asking for this for the past 4 years LOL AMD thinks they are so smart and know best.. They have proven time and time again they know jack all..They even released a 220W CPU and is hardly faster then a 5 year old 45nm 6 core Chip "they" created that runs at 125W

AMD Phenom II on a smaller 28nm 2014 CPU Node would blow away anything AMD has on the market today!

 

Yeah, it's so simple, why didn't AMD think of that? They've only been fabricating semiconductors for 45 years, what would they know?

The Phenom architecture was end of life, it was at the end of a 10 year stretch from K7 onwards, and doing a die-shrink is nowhere near as simple as you seem to think. They had to replace it with something, and a combination of factors contributed to the failure of Bulldozer.

The main problem is, once you've committed to something like changing the entire architecture that all your future derivative CPU's are based on, if you choose wrong there's not a lot you can do about it in the medium term, because CPU architectures are complicated (to say the least) and take a very long time to conceptualise, design, and verify.

Choosing a new architecture which will enable iterative improvements as manufacturing processes mature and then scale down is incredibly speculative work.

On top of that they had contracts with GlobalFoundries that forced them to buy a specific amount of fab capacity, whether or not they needed it, which drained cash that could have been spent on R&D. This was exacerbated by GlobalFoundries failing to hit timelines which delayed their already sub-par releases even further.

You also appear to be ignorant of the work they're doing with heterogenous compute. No other company has APU's that can do what AMD's are capable of, in terms of memory sharing and context switching between CPU and GPU, as well as the audio capabilities. There is a reason why Sony and Microsoft both chose AMD for their latest consoles, and it's not because "they know jack all". Intel and Nvidia simply haven't invested the resources into this area, there is no competition.

Lastly, Bobcat and it's derivatives are a much more successful architecture, managing to compete with Intel's chips in the low power segments whilst being a full node behind on process tech, which is pretty impressive.

Bulldozer was obviously not a success, but that is not a reason to assume that their next architecture will fail. Getting people like Jim Keller and Raja Kaduri back (they are both ex-AMD/ATI) can't really hurt, they have proven track records.

 

I just assumed it was a cost issue. AMD was able to provide an adequate solution at a cheaper cost than competitors.

 

Well, yes. But AMD were able to offer a cheaper solution because they're the only company that's got a really good integrated CPU and GPU.

The alternative would have been an Intel CPU and an nVidia GPU. The chips alone would probably cost more (having to buy two chips from two suppliers instead of one chip from one supplier). Then there's more RAM (so the CPU and GPU can have their own RAM), a much more complicated PCB (to handle two chips and the extra RAM, plus communications for all of them), a larger heatsink (to cover two chips), higher failure rates in manufacturing (due to the higher complexity), and so on. Plus the initial lawyers (I'd imagine that contract negotiations for a many-million-dollar project are not cheap) and engineering (persuading the Intel CPU and nVidia GPU to play nice).

Or they could just buy an AMD chip. One chip, plus a bit of RAM. Simple PCB, AMD probably handed them a reference design for the PCB to save time there, a single chip to keep cool, etc.

 

Wonder what socket they will be?

 

Unfortunately

AMD has a proven track record of what they know !
And it only has one or two shinning stars , the rest is
And please don't come back with the budget stuff ..

Often it is hard to see what is right under your nose , and even worse for fanboys to see it first and call for it - now that would make AMD look right silly would it not ..

Give the consumer what they want - oh and wait - maybe be a intel killer ! Geez Louise , you cant have that !

And the simple fact is , that's what intel did .. Slap two dual cores together and viola , a AMD killer .. It was so simple , so silly an idea , it almost bankrupted AMD . So why is it crazy or nuts to put two quads together for a 8 core , shrink the die as much as possible , and maybe create a decent CPU .

Ok so maybe it wont be a true 8 core , maybe the MB and OS will see 2x quads , does anyone really care ? As long as it pumps !
How about 4x Dual cores .. OS and MB see 4 dual core CPU's .. ????
That could Pump some life back into AMD .. As well maybe some more FPS in games ..

Me thinks AMD is short on imagination , and even shorter on delivering .
My CPU is so freaking old now , its the oldest CPU I have ever had .
Im supposed to wait for AMD to come to the party how many more years ...

I have money for a new puter , I want some power NOW! I have waited and waited and waited and waited and IM getting older , my hair has grey in it .

Seriously , AMD can get stuffed .

A fanboy no more ...

 

AMD gets torn to shreds on IPC, what you are suggesting is what they are doing already. Brute-forcing performance by adding more cores or higher clockspeed requires more power and you end up with products like the FX 9590 which is pretty consistently beaten by Intel's 4770K while the Intel uses about 1/3 of the power (Link). They need to improve their IPC but Intel has billions of dollars more to put into R&D so AMD looks likely to never catch up. It's a shame, but AMD can hardly be said to be "not trying".

 

Well

The power thing , why you shrink cores ..
I know its hard to shrink a 8 core for AMD , to get the power levels down ..
Would it not be easier to shrink a quad , or a dual core die ..
And rather than trying to make a 8 core cpu that's a true 8 core ( CHEAT ) like intel did ..

Was it the Q6600 that was like all the rage for so so long .
I built one for a friend , and for years it just roosted me for Video encoding ..
What took 10 minutes for the Q6600 took half an hour for my top of the range AMD at the time .. And things have not changed ..

I built my friend a i5 system , and it just kills my AMD in almost everything , and that's at stock clocks VS my 1090 @ 4GHZ .. It hasn't been a contest for the longest time ..

Oh well , I don't know if I want to buy another CPU , the 6 core is cheap enough , the 8 core runs 4ghz stock .. My 1090 is just about over the hill having being pushed hard for so long ..

I don't know , maybe a 8 core ( Last one ) and I might grind the heat spreader off the 1090 and see if a custom fitted head spreader does something for it .. One last hurrah , see if I cant push the 1090 past 5ghz on air .. But it is getting tired , 4 years @ 4ghz , Ive never had a 4 year old CPU before .. Its survived 3 mother boards and 2 sets of ram ...

I thought I would be doing 5GHZ + on air by now ...

 

Your system has issues, i have used both and come away with a bitter taste with intels current range of 1150 CPUs. BS costly for stuff all. When you see real world gaming results etc there is no real difference between intel and AMD esp at 4k.

Also intel fanboys are so funny.

 

Not as funny as AMD fanboys!

 

what about apple lol.

AMD are still a whole node behind and can still sell chips that compete with intel, just not with benchmarks

It is hard for humans not to be fanboys tho. Its the nature of the beast.

 

my 8350 stuck at 4ghz because of power issues (suspect mobo) does not break a sweat in any game I have tried so far. It can encode a dvd in 10mins too

 

Not a chance.

That's actually not what happened.. Dual core Conroe was the AMD Killer. Quad cores were niche when the Q series was initially released.

AMD's mistakes were made way back when K8 was dominating. They rested on their basic architecture too much and were caught off guard by what was at the time a revolutionary Uarch (conroe), launching on a very successful (I'd say unexpectedly so ) 65nm process.

It would have been no walk in the park engineering that chip.. It's roots still heavily entrenched in Intel's current Haswell processors some 8.5 yrs later.

K10 was already well into development at that point, and about hit the first of several delays due to a 65nm process that was frankly, terrible, as was evidenced by the K8 shrink.
Thefroe it was a combination of the K10 architecture missing the mark next to conroe both in IPC (around 10% shortfall) and frequency (impossible to say how much by, but not much) , on a massive failure of a process that meant it launched at a ridiculously low frequency.

45nm was a success in some ways and showed what K10 was capable of by taking away that process frequency wall, but was late, and so on it goes.. an impossible game of catchup from then on. There was a tiny window of opportunity that was only just missed with this shrink which would have made it competitive in the top end for a small time, but from then on there was no making up ground.

All this was not to say AMD couldn't have done a lot better, but the damage was done way back then, and I think even with better management and direction over the last 8yrs or so, there still would have been NO hope of an actual "return to glory" Intel would never let it happen. You just cannot develop killer chips without resources, and AMD have never really had any