$500!==??!??! http://wccftech.com/amd-zen-8-core- 35ghz-cpu

$150!!!

BASED AMD DID IT

>>57684179
Hahaha poorfags btfo

I told you fags weeks ago a binned chip would be $500, absolutely no one responded.

AMD should rot away, Nvidia is 10 times more bigger

http://finance.yahoo.com/chart/AMD#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

>>57684179
Look at the projected cinebench benchmarks. The CPU is competing with a $1000 CPU.

If you want cheap, go with the 4/8 core CPU for $150 that is competing with Intels i3/i5/i7

File: grand wizard.png (400KB, 816x1020px) Image search: [Google]

400KB, 816x1020px

>>57684179
Buy it now.

You will thank me later.

>>57684286
Delete this

>$150 4c8t (i7 equivalent)
>$250 6c12t
>$350 8c16t
>$450 8c16t

Intel's 8c16t is $1000+
Intel's 6c16t is $430+

So half the price, not so bad.

>>57684286
Its market cap is nowhere close to being 10 times larger and Nvidia has a higher stock price than Intel and Qualcomm combined, what do you think that means?

How am I supposed to respond when you clowns bitch when AMD makes itself look like a 'poorfag' brand and then flip your shit when it puts out a premium product?

>>57684179
Im waiting until the performance benchmarks come out. Pricing is irrelevant until then.

>>57684179
>headline is $500 Zen chip beats $1000 Intel chip
>ohmahgad so expensive

even if we forget your utter retardation those are still leaks and should be disregarded

Intel's EX line is 130W, isn't it? How can a 95W part compete even if it has the same number of threads?

Is the 130W TDP on Intel under AVX loads?

>>57684370
Intel's TDP calculations are ultra retarded, you get situations like 15W ULVs using around 40 watts.

Their ratings get less silly when you approach HPC Xeon wattage levels, since don't fuck with those, but for consumer crap their TDP numbers are pure nonsense.

>>57684370
>AVX
Absolutely no one is using avx right now, friendo
(Knights ridge is the only one)

>>57684448
Haha great meme friend

>>57684448
Shit do use AVX if you compile your shit with the right USE flags.

>NEW GENERATION BEATS OLD GENERATION

STOP THE PRESSES!!
AMD HAS INVENTED HOT WATER, HOLY SHIT


Retarded stupid AMD shills

>>57684179
500$ is binned chip you faggot

>>57684527
Specifically, what cpu do you have that uses avx?

>>57684544
You can cry all you want until Skylake-EX comes out, somewhere in Q4 2017.
Meanwhile, Zen is releasing in about 6-7 weeks.

>>57684585
Oh wow, so they are finally catching up to a 5820 after all this time? And it only takes 8 cores for amd?

File: lolsuika.png (322KB, 729x717px) Image search: [Google]

322KB, 729x717px

>>57684179
AYYYYMD POOJEETS ON SUICIDE WATCH
>muh $300 octa core

>>57684581
On the machine from where I am writing this post a Xeon v3, which supports both AVX and AVX2.

>>57684179
hahahahaha

so much for being the affordable line up

I CAN BUY A $100 IVY BRIDGE 8 CORE FUCK OFF AMD FUCK OFFFFF

FUCK YOU AMD FUCK YOU PIECES OF POO IN LOO SHIT ZEN IS A FLOP ZEN IS A FLOP ZEN IS A FLOP ZEN IS A FLOP !!!!!

>>57684624
But not 512

>>57684652
>>57684667
>false flagging poojeet

>>57684689
Go away /v/, and take your idiot circus freak with you.

>>57684688
And your point is exactly?

intel fags on suicide watch

>>57684743
Haha retarded AMPOOR(IAMPOOR) shill Intel has Zen level performance for years already get a clue retard shill

>>57684743
Actually I am looking forward to replace my aging Xeon v3 with one of those bad motherfuckers from AMD if they prove themselves worth something.

so zen has haswell ipc?

>there are retards ITT right now who thought amd would offer true octa core for less than $500
Bulldozer isn't octa core and that's why it's so cheap, retards.

>>57684736
What was your point in bringing up avx in the first place?

>>57684802

IPC-wise, we don't know. That is the bare minimum at least to expect but I think we can expect between Broadwell and Skylake levels of IPC.

>>57684448

That's where you're wrong. Off the top of my head, Blender will use AVX if you have it available for rendering and Prime95 started using AVX since version 27. FMA is not something you will pass up if avalible on computationally expensive workloads which is why Intel has been adding it to their chips since the Sandy Bridge days.

>>57684824
They're offering 8c16t for $350, if the leak is true.

>>57685014
>>57685014
>Broadwell and Skylake levels of IPC
AHHAAHAHHAHAHAHHAAHHAHAHAH
fucking shill

>>57684179
Compared to the 1000$ for a comparable Intel chip its a goddamn steal.

>>57684824
There's a model for $350 usd with slightly lower clock speeds

At 500 USD buying opterons will probably make more sense...

>$450
>Lower single-thread performance than a Kabylake Pentium

>>57685260
Pentiums can clock high as fuck because low coar count

I want to see what frequencies a dual core zen can reach

>>57685014
512 it should be implied avx is being used since everything has that.

>>57684179
>half the price of similar benched intel processor
>still complaining
Poor people remain poor.

>>57684179
Intel and nvidia will go bankrupt

>>57684179
> I THOUGHT ZEN WAS SUPPOSED TO BE CHEAPER

"Zen" is not a line of chips, it is a micro architecture.
Summit Ridge is the consumer CPU line.
Lisa Su said in an investor meeting that their upcoming parts would have a higher ASP than previous FX parts.
This is after she gave a little speech saying that AMD was done being the budget alternative.

Summit Ridge is aimed at competing with the i7E line of prosumer chips from intel.
AMD's mainstream competitor to the higher clocked i5/i7 SKUs is their upcoming Raven Ridge APU line.

>>57684802
IPC isn't a single thing, it depends on the instructions you're processing.
Clock for clock we know that a Zen core is roughly equal when rendering a scene in Blender.
Leaked SiSoft Sandra results show pretty decent performance as well.

What we know is that AMD touted average IPC to be a 40% uplift over Excavator, and that puts it all over the place. Performance in certain metrics could only be on par with Sandy Bridge, while its clearly quite higher in other areas as shown in the Blender comparison.


>>57684370
The i7E parts tend to draw a bit less than their TDP under normal load. They'll usually hover around 110-130w. In the newer chips they have an entirely different boost state with separate clocks specifically for large AVX instructions. The data paths ways pull so much power that you can actually see them physically switch on in a thermal image.

Zen cores are not as big as intel's Core i arch, they flat out have less transistors. Zen won't be competitive with intel's arch across the board in everything as a result. Thats how they're getting away with an 8core/16thread chip that only has a 95w TDP.

>make expensive chip
FUCKING FAGGOTS NOBODY WILL BUY IT
>make poorfag chip
FUCKING FAGGOT POORFAGS

AYYMD IS FINISHED & BANKRUPT

>>57685146
Haswell -> Broadwell -> Skylake
Count up all of them and you get around a grand total of 12% IPC improvement in non-synthetic integer loads.

>>57686097
you mean 6%?

>>57686183
I'm being generous.

>>57684313
But it's less than half the performance due to the sub 3.0GHz clock speed.
D
O
AMD

>>57686261
This, ZEN will have 2.5GHz base and 3.0 Boost if you knew anything about CPUs you'd know this fact

>>57686261
>>57686356
Cute shitposts, children.

>>57686404
Even the experts at realworldtech think so, tech illiterate retard

>>57686447
Summit Ridge is clocked over 3ghz with the same energy required per clock as Excavator.
Again: Cute shitpost.

>>57686461
Kek, no.
That's completely and ignorantly wrong.

>>57686526
When AMD showed off their 3ghz Summit Ridge competing with and out performing 3ghz Broadwell-E they told the press that final silicon would have higher clocks.
A few weeks later another ES surfaced online with a 3.15ghz base clock, and even higher turbo.

These are objective facts.
So yet again: Cute shitpost.

File: drop_the_mic.jpg (20KB, 701x410px) Image search: [Google]

20KB, 701x410px

>4core/8 thread Zen is 150 bucks
Thank you based Keller

>>57686576
certified shitwrecker

File: 1272976906843.jpg (36KB, 370x300px) Image search: [Google]

36KB, 370x300px

>>57686559

File: certified shit wrecker.jpg (13KB, 381x286px) Image search: [Google]

13KB, 381x286px

>>57686604
Goddamn right.

8C16T - 3.2/3.5 - 95W
6C12T - 3.4/3.8 - 95W
4C8T - 3.6/4.0 - 65W

I hope i can undervolt these to hell.

>>57686686
With a Zen core pulling 10w~ for a stable 3-3.2ghz clock, they should pull only 7w for 2.8ghz following Excavator's clock scaling.
A 95w chip will turn into a 65w chip without a sweat. They should scale down even further than that.

The 4 die Opteron MCM will probably be clocked around 2.2-2.5ghz, and that power envelope is just 3w to 5w per core.
Consumer Summit Ridge should make for a solid competitor against intel's Xeon-D family, and their socketed instead of being soldered.

>>57684273
I remember anon. AMD needs this though. People just want them to be bargain-bin cheap shit again. Zen is good.


>>57684179
>>57684619
>>57684651
>>57684652
>>57684667
>>57684764
>>57686083
>>57686261
>>57686447
There are 5 more expensive processors from Intel on that list. Eat shit and die you Intel jewish defence force.

>>57685146
This
Zens competition is the low clocked xeons with a dozen cores

I'm way more interested in 4 core APUs for mobile if IPC gets close to Haswell.
Because then I won't have to deal with Intel's garbage 3D drivers and even worse performance.

are these any better than a 2500k or is this just some multi-core budget trash they hand out to Disney animators in Korea ?

>>57686447
There are samples out with 3.2 Ghz clocks.

File: 1478825100806.jpg (264KB, 750x1334px) Image search: [Google]

264KB, 750x1334px

>>57686639
>8 core
>3.2 base clock
>he actually believes this
Kek

Why are you niggers assuming clockspeed when you don't even know how wide the core is, how many stages the pipeline is nor do you know what process GloFo is using.

Are you fucking retarded.

File: 7-1080.1262858105.png (101KB, 1920x1080px) Image search: [Google]

101KB, 1920x1080px

>>57687019
>Why are you niggers assuming clockspeed
We've seen engineering samples, and AMD themselves explicitly said they'd have chips clocked higher than 3ghz.

>you don't even know how wide the core is,
We have known how wide the core is for a year. Do you live under a fucking rock? Rhetorical question.
AMD detailed the arch at HotChips too which put to rest any doubt.

> nor do you know what process GloFo is using.
Its Samsung and GloFos shared 14nm LPP.

Why did you even make this post when you're clearly the most ignorant one here?

>>57687081
How much instructions and retires can a zen core do per clock cycle?

>>57687150
Listen to the HotChips presentation.
https://www.youtube.com/watch?v=sT1fEohOOQ0

>>57687187
Watch this ad
I'll just wait for benchmarks to come out and the inevitable excuse of why zen isn't as good as "between skylake and broadwell-E" in single thread like /g/ promised

>>57687150
http://www.anandtech.com/show/10591/amd-zen-microarchiture-part-2-extracting-instructionlevel-parallelism/7

There's a table there.

>>57687207
>a technical presentation is an advertisement!

Beg for information then turn your nose up at exactly what you asked for.
You're a butthurt retard.

>zen isn't as good as "between skylake and broadwell-E" in single thread like /g/ promised
No one but retards on par with you have made these claims.

>>57687207
You have to be some kind of mental patient to listen seriously to what /g/ tells you.

>>57687207
Judging by the "better than 40% IPC increase" it's around Haswell levels of IPC. Even if ends up slightly lower clocked than Kaby Lake it will completely blow away anything Intel has because you get an 8 core Zen for the price of a 4 core Intel.

>>57687316
There's 3 architectures after Haswell, each one improves performance by at least 10%, do the math.

File: (JPEG Image, 240 × 210 pixels).jpg (8KB, 240x210px) Image search: [Google]

8KB, 240x210px

>>57687014
Why not?

That would be the same as the 6900k

the only thing I'm having trouble believe is that TDP

>anything sub 4Ghz
Not sure what you kids are trying to do but it's kinda cute.

>>57687423
What are you talking about you retarded fuck.

>>57687351
the simple math is that a 3GHz 8 core Haswell is currently selling for $1000 and completely blows away any Skylake or Kaby Lake quadcore.
Zen is going to sell the exact same thing for half the price

>>57687351
Broadwell was a 2-3% uplift over Haswell in most ops.
Skylake was again about 3% uplift over Broadwell.

Intel has not been providing 10% performance uplifts generation over generation unless you exclusively looked at synthetic specific FPU ops.

>>57687372
1D3201A2M88F3_35/32_N

Valid ID string from Summit Ridge samples that have been sent out to OEMs.

It 3.5ghz turbo, 3.2ghz base clock, and its still 95w TDP.
They have a relatively low TDP for the core count because the cores aren't as big as intel's. AMD's Zen won't be competitive across the board vs intel's cores.
They're smaller, lower transistor count, and don't have nearly as big of an FPU. They don't have native 256bit data paths at all.
Thats why they can manage the lower TDP.

>>57687782
We can expect a new turbo from AMD, but how is it gonna work? It says 3.5, but for what? Only a single core? Two cores?
Is there any chance of going higher than the specified turbo in some lightly threaded single core workloads like Intel does?

This is pretty important.

>>57687867
They emphasized highly articulate gating of unused hardware resources to save power. Socket AM4 also has the strictest tolerances on power delivery of any thing AMD has ever put out there.
I assume they'll allow a single CCX to be active in an 8 core chip in more serial workloads. Gating off a full CCX and allowing 1-4 cores to clock higher and take full advantage of that remaining thermal margin.

They haven't gone over any exact details on their boost states yet, though they are already doing this in Excavator which can shut down a module as well as part of its L2 to save power.

Worth mentioning is that idle power draw is supposed to be very low.

>>57687867
>Is there any chance of going higher than the specified turbo in some lightly threaded single core workloads like Intel does?

This is only a feature of Turbo Core 3.0, which Kaby Lake lacks for whatever reason.

I have a feeling that the 500$ model is going to perform nearly exactly like the 6800k, not like it's equivalent in the core count, the 6900k.

It would be great to see these guys wreck Intel and to push them to compete, but AMD has given me little reason to believe that they could pull that off.
Hope the certified shitwrecker did his job properly and performed a miracle.

The unlocked i3's and 6 core mainstream parts on Intel's roadmap pretty much show Intel doesn't consider Zen a joke like the Bulldozer derivatives.

>>57684179
go fuck yourself, we want a competitor to intel, even if it means higher than poorfag prices

amd almost got fucked because it only made cheap turds to attend to poor people instead of focusing on performance

>>57687985
It's not just about the high end. Intel has completely neglegted the lower end of the market. No improvement since Sandy Bridge.
That's where Zen is really going to wreck shit.

It's still a MCM design, the more cores it stacks up the more they'll fall behind monolithic Intel dies due to inteconnect overhead and coherency.

This has me really worried.

>>57688235
Anon, go home. You're drunk.
Consumer Summit Ridge is a single die, only the Opterons have an MCM package.
You have literally no idea what sort of interconnect fabric they have. They aren't using a derivative of HypterTransports, its entirely new IP, and they have stated that its fully coherent.

Concern trolling is the shittiest form of shitposting.

>unlocked 4C/8T for $150

Shieeet, even if the IPC is only around Sandy Bridge levels, that's going to be a great chip for poorfags.

>>57688278
I'm not worried about Summit Ridge, I'm worried about Naples.

>>57688278
Also you're right about me being drunk, I played Dead Space drunk on homemade wine.
Holy shit what a blast.

File: 1478538517012-pol.gif (966KB, 312x213px) Image search: [Google]

966KB, 312x213px

>>57684179
Why are you upset. The literal 4core 8thread (i7 haswell equivalent (minus clock preformance)) for $150!!!
The $500 one will likely have huge cache and clocked like a god. Dude all of amd's (desktop) chips are unlocked. What is the problem with this???

>>57688281
Exactly. A haswell i7 unlocked with updated features for $150. It's fucking happening!!!!

Fucking bump

File: Zen team.jpg (75KB, 715x573px) Image search: [Google]

75KB, 715x573px

Based

The next generation 7nm Zen+ parts will be even more amazing.

>>57688544
I'm thinking, what about that Intel shenanigans with compilers? It's still happening? AMD CPU's may come with all new instructions, if programs don't make use then it means nothing.

>>57684349
This.

>>57688716
>what about that Intel shenanigans with compilers?
ICC is still technically updated on a per processor basis, but its capable of putting out high performance binaries for any of AMD's architectures. How well a program performs for multi core chips is entirely up to how it was coded in the first place. There is no aberrant performance disparity between ICC and GCC when you compare them side by side.
Of course this could easily change at any time. intel should never be trusted to be honest.

> AMD CPU's may come with all new instructions
They don't have any new proprietary instruction sets. They dropped a few introduced for the Bulldozer family.

>>57684179
www.reddit.com/r/Amd/comments/5dl2ho/what_amd_is_up_to_with_zen_disscussion/

the AMD specs are made out of rumors and some leaked facts. So don't take any of that as facts set in stone.

>>57689157
>using reddit at all
>cross posting reddit shit here

File: 1469877426690.png (124KB, 358x272px) Image search: [Google]

124KB, 358x272px

>>57684179
>Zen projected somewhere between Ivy and Broadwell IPC
>Somehow it's shit if you can't get it for 1/4th of the price

>>57689202
At least they don't have Pooloos shitting up their threads. Hardware on /g/ is literally shit now. It's gotten so terrible I could get better discussions on /v/ sometimes.

>>57690499
Its only bad because no one pushes against the low brow shitposting that has become commonplace.
/o/ works pretty well because if you don't know what you're talking about you'll get called out instantly. You can talk out of your ass on any topic here because the average poster doesn't know a damn thing about any relevant topic.

File: 1479974300011.jpg (126KB, 600x720px) Image search: [Google]

126KB, 600x720px

>>57686261
>comparing clock speeds from two different architectures

>>57684179
Here is the sticking point.

notice the 8/16 right below it that is 350$, this is where im wondering, if this leak is real, what is the difference between the 350 and 500$ cpu.

>>57684349
we want the premium product, but we don't believe that the price of currently premiums are worth it. Especially if this 500$ cpu is all just binning and nothing more.

>>57686056
Amd is in a weird space
If they make their shit cost to much, you put it against intel and why get amd if intel is in your budget you are willing to pay.

if they make it too cheap, they make no money. they have a weird balance they need to strike.

>>57691155
> Especially if this 500$ cpu is all just binning and nothing more.

What the fuck do you think SKU differentiation is, nigger?
Protip: Its all entirely defect and clock binning.

>>57691271
Im assuming that all the chips are based on the same lithography right now, as in amd is making 1 cpu and through mcm and cutting gets skus.

the 4 core is a fucked 8 core, or possibly a cut in half low bin 8 core, the 6 core has 2 fucked cores and the 8 core is everything works.

but that 150$ up sell from 8 to 8 special... that is something I want to know what the fuck is under the hood far more then anything else. Is the clock increase that significant? are they that overclockable that this out could go passed 4? really this is the sticking point currently.

we wont know till they land, but im looking at that 350 and thinking that's my cpu unless the special is special for a god damn reason.

>>57687985
i want to see mixed workloads honestly, because we seen one workload where amd beat out intel 8 core at same clocks. i'm really interested in knowing what large fpu workloads are, and how often they are used, because personally, If cutting the really heavy shit out that is rarely used just makes the heavy shit a bit slower, but equal in all other loads... im 100% ok with that.

>>57688834
>>57688716
the compiler shit was handled a long time ago, however those tainted compilers are still used so yea...

>>57691314
>the 4 core is a fucked 8 core, or possibly a cut in half low bin 8 core, the 6 core has 2 fucked cores and the 8 core is everything works.

This is standard practice for literally everyone who makes anything in silicon.
6 core i7Es from intel have binned off cores.
Now with Broadwell-E being a full 10 core die, the 8 core models have binned off cores as well. The i7 6900k is a $1,000 partially defective die.

Are you so new that you think AMD is the only one doing this? Really?

>>57684565
What did he mean by this?

>>57684179
It's SPECIAL.

>>57691408
They need some way to rationalize that Intel will still be making the "superior" processors.

I'm ready for a new build and I can't wait for these chips to drop.

God damnit I want some real fucking benches released. How high can these bad boys be oc'd?

>>57691208
>can't price too high because of competition
>can't price too low because they're trying to make money

No weird would be pricing a product in a situation where those things weren't a fact of life. That's standard shit.

>>57684286
nigga i ain't clickin that

>>57691314
Isnt this + pcie lanes whats hsppening with the 6800k and 6850k?

If amd is currently competing in gpus upto about the 1060 and ignoring yhe 1070 and 1080...would this be the same strategy? With their 16 cores coming out at a later date competing the current intel 10core?

Honest questions from a retard

>>57692711
You're confused as fuck.
There are no 16 core consumer chips.

>>57692731
bruh....

>>57692731
Yeah I probably am. But not many consumers are buying the 10 core chips are they? I have honestly never owned an amd cpu did opterons allow overclocking? Will we see something like what happened with the (nehalem I believe?) Dual/single socket solutions for people in that budget range?

Im just trying to understand how as a consumer everything fits together as I want to upgrade from my 920. Ill wait for reviews/benchmarks since im not in a huge rush but trying to stay somewhay informed even if all the technical talk flies way over my head.

It was supposed to be 700 so get your panties out of a knot

I am so grateful that I did not sell my AMD shares when they hit $4.00

>>57692761
Fuck off with your retarded shitposting.


>>57692763
> But not many consumers are buying the 10 core chips are they?
They're not really targeting the typical consumer market, these are what is referred to as "prosumer."
They're basically workstation chips one tiny step below a Xeon. Selling them at $1,700 a pop they really don't need to move a ton of volume either.

I'm not sure if Opterons have a locked multiplier or not, I've never bothered checking. I generally wouldn't ever risk compromising reliability on something that needs 100% uptime. As for dual socket consumer AM4 boards? I doubt that novelty will emerge again.

>>57684179
>4c8t
>$150

That's pretty much an i7 3770, but slightly cheaper than an i5 6400.

>>57692876
More like an i7 3770K

>>57684179
>8core CPU
>twice cheaper than intel
>complain

fuck off

>>57687014
The earlier 8C16T ES's base clock was 3 Ghz

>>57686261
>what is overclocking?

>>57684273
acknowledged.

Now let's wait for real world performance benchmark

>>57684273
>I told you fags weeks ago a binned chip would be $500, absolutely no one responded.
>binned
>8c/16t
What? What about this do you think is binned? Are you mistaking this for the MCM 16c/32t package?

I will buy the $500 Zen.

File: 1476275762826.jpg (218KB, 512x800px) Image search: [Google]

218KB, 512x800px

Reminder that /g/ will judge AMD Zen by its ability to run non-free proprietary video games for Microsoft Windows, as if that was the one universal metric to judge CPUs by.

File: 1452274458515.jpg (92KB, 768x783px) Image search: [Google]

92KB, 768x783px

GO AMD GO

>>57694253
99.99999999999% of posters here don't understand what binning is in any context.

The top end Summit Ridge is a "binned" Zeppelin die. It is the highest binned consumer chip. It is binned for the highest target clocks of any consumer SKU. Just saying a chip is binned doesn't mean anything, all chips are binned. Thats like saying you were a graded student in school. All students are graded, and not all grades are the same. Without context the statement carries no meaning.

Its like people literally just learned this term and don't know how to use it. Just like with retards throwing around the term bottleneck.

>>57694389
Binned in this context means 'golden sample', aka the 3 in 150+ dies per wafer.

>>57694145
>AMD
>overclocking
nice joke

i5 6600 4c/4t $220

zen sr5 6c/12t $250

fuck yeah, sign me the fuck up, probably gonna get the $350 8c/16t

>>57684299
>>57686617
Is there anything Jim Keller(pbuh) can't do?

>>57684313
> cores equal performance
WEW LAD

File: shit wrecker.jpg (62KB, 800x400px) Image search: [Google]

62KB, 800x400px

>>57694620
He can't not look like he could probably be Kevin Sorbo's father.

how will these compare to skylake-x?

When do you think will be the next iteration?

>>57694693
There will probably be another 14nm refresh in 2018, same arch just higher clocks.
Zen+ chips will be 7nm parts coming 2019.

>>57694024
That's not a bad thing though, why is op so upset?

That's still 25% cheaper than an i5 6400(australian prices)

>>57684313
This
>>57686261
Don't forget they all have unlocked multiplier even the $150 4c8t one. Which intel now releases 2c4t (i3 7350k) with unlocked multiplier for $177.

>>57694644
clock/clock they're same according to Cinebench

ofcourse not all programs are same, but with what little basis we have, thats the loose assumption

>>57694764
With equal clocks Summit Ridge ended up finishing a small scene about 2% faster than 8 core Broadwell-E.
You could argue that AMD had the benefit of compiling Blender specifically for that demo, but its never made much of a difference in any independent testing done.

We can guess that it'll be pretty decent in these type of rendering workloads.

File: wccftech.jpg (52KB, 400x400px) Image search: [Google]

52KB, 400x400px

>>57684179

>>57695071
/thread

>>57684179
AMD is total shit, one of the worst tech companies ever right now.

They need to die completely.

>>57695533
Top bait, my man. Well meme'd.

>>57695800
How does it feel Zen is a total disaster?

>>57695811
Its feeling pretty good that you're so distraught its causing you to shitpost.

>>57695811
idk, I guess you would know better what a total disaster feels like

>>57695811
You could at least wait until it's out before saying it's a disaster.

>>57684179
Maybe if it proves to be a good OC-er i might have a reason to Upgrade from my 4790k i'll be waiting for 3-4 months after release.

>>57684846
You did, I never mentioned AVX before you did.

How much will the APUs cost

>>57695838
shots fired

Should I go with a skylake i5 or wait for zen?Im running and old phenom ii 965 now.

>>57697464
Always wait for the new gen when it is this close to release

>>57684179
I don't get it. Isn't this thing supposed to compete wity intel's $1,100 8-core haswell i7-5960X?

$500 for a ~4GHz 8-core zen processor seems like a pretty gud deal to me.

>>57684179
>tfw cheapest Zen costs more than their high end FX-8350

AMD is confident in their product.

Why are cinebench scores for Zen 2 times higher compared to same priced Intel processors? Why aren't the pajeets on /g/ talking about this in the first place?

>>57684619
>"special" SR7 costs $500
>regular SR7 costs $350
The special one probably has SMT the regular doesn't. The price leaks implied >$300 which turned out to be true.

>>57684179
>I THOUGHT ZEN WAS SUPPOSED TO BE CHEAPER
but it is, compared to intel

>>57697613
Estimated Cinebench scores from an unkown source aren't a good basis for discussion

>>57685260
Is this bait. The highend i3 and pentiums at 4ghz which means their single core performance is similar to i7.

>>57697749
This whole thread is about the "unnamed" pricing. If your argument would've been a thing this thread would've died much sooner.

File: 850px-broadwell_block_diagram.svg.png (114KB, 850x1075px) Image search: [Google]

114KB, 850x1075px

>>57687081
Zen is basically AMD's broadwell. They will deliver.

>>57697851
I'd rather it be another Athlon 64.
More likely though it's another bulldozer.
I mean come on- it's fucking AMD. There's going to be SOMETHING wrong with it.

>>57692143
http://finance.yahoo.com/chart/AMD

>>57688716
Well considering intel got caught and sued last time i imagine its nowhere near as crippling or a non issue entirely now

File: 1471938619850.jpg (23KB, 232x334px) Image search: [Google]

23KB, 232x334px

>>57684179
ayyyyyyyy

>>57691408
I dont think he mentioned it being unique to AMD. In fact, based on his assumptions it means he knows its a common practice.

Its no different with gpus etc. Remember news of flashing a 4gb 480 into an 8gb 480? Same idea

>>57694563
So just say the highest binned chipped

>>57698137
Chips*

damn phone

>>57697603
Yeah but if I say that I won't get my 2 rupees.

>>57697603
AMD cores aren't the same as Intel cores. That'll be on par with a 4 core i7 at best.

>>57698413
>Zen
>A complete redisigned cpu architecture with 1 FPU per core and overall 40% better IPC over excavator
>this is going to have the same performance as a 5GHz FX 8-core bulldozer processor
lol k

>>57699118
But it doesn't have the magic intel pixie dust that makes the cores faster.