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

[TheRagingGamer]

dude chill.. you're taking what I've said out of context. and for the note neither have you so... moot point

yes neither have i but 14nm>22 nm

[harrynowl]

yes neither have i but 14nm>22 nm

By 2016 which is when we expect to see this 'Zen' Intel will be rolling out 10nm.

 

Process size =/= performance (it correlates to potential performance though). Otherwise my 45nm Core2Duo would not beat my 32nm A8 5600K in single threaded tasks.

[TheRagingGamer]

By 2016 which is when we expect to see this 'Zen' Intel will be rolling out 10nm.

 

Process size =/= performance (it correlates to potential performance though). Otherwise my 45nm Core2Duo would not beat my 32nm A8 5600K in single threaded tasks.

but that's 2016 this is 2015 

[harrynowl]

but that's 2016 this is 2015 

Intel will have 14nm in 2015 too (remember production=/=release, don't expect amd to have these cpus out by 2015). AMD don't have a fab or money to play process shrink games with Intel, this has been true for over 10 years now even when AMD had the perf crown they were usually behind. I'm just looking at facts here. Like you I have no idea how performant these new CPUs will be but based on facts it'll be faster than FX but probably not an Intel killer. Cant compete with Intel R&D budget unless there's a massive screwup like P68.

 

90nm: dothan/prescott (intel) 2004 Q1

90nm: winchester (amd) 2004 Q4

 

65nm: cedar mill (intel) 2006 Q1.

65nm: brisbane (amd) 2007 Q2.

 

45nm: core (intel) 2007 q4

45nm: k10.5 (amd) 2008 q4

 

32nm: nehalem (intel) 2010

32nm: zambezi (amd) 2011

 

22nm: ivy bridge (intel) 2012

28nm: steamroller (amd) 2013

 

14nm: intel 2014

14nm: amd 2015

 

10nm: intel 2016

amd??

[patrickjp93]

dont bother with him, he is so deep in the fanboy he doesnt even accept nvidia also, he claims that AMD and NV wont improve their chips in 2 years and intel iGPUs will have dominance over the midrange sector. you can see how non fanboy that is... /sarcasm

(and this was written by someone who by choice (mostly irrational) would never use AMD, and would consider himself an intel/NV fanboy at heart)

AMD and Nvidia will improve their chips certainly, but Nvidia's done unless it gets a helping hand from Samsung, and yes, I'm pretty sure Intel's iGPUs will be stronger than Carrizo when Skylake rolls out, leaving Intel holding most of the low and mid-range markets.

[patrickjp93]

I'm not so sure about that.

However if I remember correctly westmere have 16 pipeline stages, where haswell only have 14 pipeline stages.

More pipeline stages will increase the CPI (loop detection, prediction is used to shorten the pipeline).

If we don't take things like die-budget and heat into consideration, then yes, frequency does not degrease the IPC.

However because we are limited to a die-budget and a theoretical TDP, we cannot have both.

IPC requires alot of diespace. Increasing heat and powerusage means you will have to run at lower frequencies.

My example was about software, not hardware. Multiple workers, would be considered threads.

Parallel code, the threads should never intevene or share elements with each other.

Good code will have less data races, however some algorithms are doomed to data races. Not because of bad coding, but the nature of how you potentially solve a problem.

And, it would be considered thread-level parallelism not instruction-level parallelism.

Then you really need to work on your English, because your explanation makes no sense against your example.

[vm'N]

Then you really need to work on your English, because your explanation makes no sense against your example.

An example would be having multiple workers work on the same ressourcepool.

I should have used the term "datapool" instead?

However I did use the term "worker", which really belong to the software area and not the architectural / hardware area.

[patrickjp93]

I should have used the term "datapool" instead?

However I did use the term "worker", which really belong to the software area and not the architectural / hardware area.

Workers are commonly the abstractions of ALUs in super scalar architecture.

[vm'N]

Workers are commonly the abstractions of ALUs in super scalar architecture.

You will need to explain further on this one.

[deviant88]

Guys im in shock! nerdshock.

14..een.. nan..oh..met...er. 

[Misanthrope]

dude, phenom2s were going effortlessly against Nahalems and Westmeres... they were even slightly better all in all

My current rig runs on a phenom II x2 560 for the next five days or so (i3 4150 on its way to me) And yes it's a great chip: Still holds up ok after like 6 or 7 years

[Unhelpful]

My current rig runs on a phenom II x2 560 for the next five days or so (i3 4150 on its way to me) And yes it's a great chip: Still holds up ok after like 6 or 7 years

I still have one in my dads rig. 

 

Bent about 9 pins to hell and managed to straighten them. That thing is still running today

[deviant88]

My current rig runs on a phenom II x2 560 for the next five days or so (i3 4150 on its way to me) And yes it's a great chip: Still holds up ok after like 6 or 7 years

I have a phenom ii x4 955 gathering dust for over a year,that thing was good for 2-3 years but it couldnt cope with bf3 and bf4 also old am2+/am3 platforms suck,especially the one i have 770/710 southbridge or something.

Cant wait for 2015 and 14nm cpu wars,and 14-20nm gpu wars + windows 10/dx12 madness,it kinda got boring lately nothing interesting.

[Sharkyx1]

I do not understand fanboys. I have been using nvidia GPUs since 2001 (Built my first rig this year)and I chose an r9 290x. Why? I got it for 450 bucks during the mining craze, which was better than the 780 at the time. I buy whatever meets my expectations and price point. I could care less

i was with you until there

 

this should be a banable offence/s

 

but yeah you are on point with that, and i don't think anyone disagrees, if AMD launched a better CPU that intel, anyone in the know would switch when it came time to upgrade. the problem is image, out in the real world Intel holds a better image and that isn't changing.

[miagisan]

I have a phenom ii x4 955 gathering dust for over a year,that thing was good for 2-3 years but it couldnt cope with bf3 and bf4 also old am2+/am3 platforms suck,especially the one i have 770/710 southbridge or something.

Cant wait for 2015 and 14nm cpu wars,and 14-20nm gpu wars + windows 10/dx12 madness,it kinda got boring lately nothing interesting.

on ultra at 1080p my 965 still does 50-63 fps (max i got it to with OC of 3.8ghz) on BF4 64 man server.....

[Admire]

they were ahead of intel before, what makes you think they won't be again 

 

Money and money spend on R&D.

 

I desperately hope that AMD comes up with something but it is a case of what have you done for me lately...

[MetroP]

Money and money spend on R&D.

 

I desperately hope that AMD comes up with something but it is a case of what have you done for me lately...

They've always managed to outshine AMD even with their limited resources.

[Admire]

They've always managed to outshine AMD even with their limited resources.

 

Wait, what? I didn't get that.....Amd has managed to outshine amd?

[patrickjp93]

You will need to explain further on this one.

At least the way I learned it, ALUs, or computing resources on a chip, are generally considered workers that can use a common resource(memory) pool and work on a task, and the scheduler(s) is like the foreman who oversees it all.

[RexinOridle]

Aaaaand ASUS law suit coming.

[dalekphalm]

Aaaaand ASUS law suit coming.

Want to elaborate on that one? How do you figure ASUS involvement in this?

[FakezZ]

14nm r9 390x anyone?

[patrickjp93]

14nm r9 390x anyone?

Won't happen. The architecture has to be designed for a process node. The die shrink isn't just magical scaling. Things move around. The 390x was designed for 20nm and that is where it will stay. TSMC is going to be in serious hurt soon if Nvidia decides it wants to work with GloFo on Samsung's proven tech. And there's no way AMD is going to stick with them and play as a second-rate citizen against Apple and Nvidia.

[FakezZ]

Won't happen. The architecture has to be designed for a process node. The die shrink isn't just magical scaling. Things move around. The 390x was designed for 20nm and that is where it will stay. TSMC is going to be in serious hurt soon if Nvidia decides it wants to work with GloFo on Samsung's proven tech. And there's no way AMD is going to stick with them and play as a second-rate citizen against Apple and Nvidia.

Well that is unless TSMC gets 14nm ready for mass production soon, which frankly I don't think is possible. Also for the 390x thing I said it was more of a joke than anything else

[Humbug]

What AMD is prepping according to rumors is the 380x.

With the 390x being saved up to compete with Nvidia's upcoming high end 980ti.