Zen benchmark The hype is dead

>>58153096
I'm wondering where my 2600k at 4.4 ghz would end up on that graph

>>58153096
These results are actually pretty good for a un-finished ES processor. Can not wait for final silicon to hit retail to finally replace my aging, yet still running 2500k.

File: amd1.jpg (71KB, 552x661px) Image search: [Google]

71KB, 552x661px

>>58153096
DELET THIS

>>58153167
You're going to replace your 2500k with an octa core? For what purpose

>>58153167
>unfinished
>month before retail availability
yeah sure

And for performance per watt?

ANOTHER MASSIVE DISAPPOINTMENT

>>58153132
>>58153096
These are the fakest looking benchmarks I've ever seen. The fuck is with the inconsistency in the image?

AYYMD IS FINISHED & BANKRUPT

AYYMDPOOFAGS CONFIRMED ON SUICIDE WATCH

>>58153096
We fucking told you this would happen but you AMDumbfucks built of hype anyway
fucking idiots stop it

File: 2001347995.png (132KB, 600x327px) Image search: [Google]

132KB, 600x327px

Hoping their GPUs are at least good

>>58153096
s o u r c e

File: 2ec.png (572KB, 600x580px) Image search: [Google]

572KB, 600x580px

>>58153318
F-F-F-Fake

>>58153341
This, where is the source?

8 fake cores slower than Intel quad cores

>>58153342
post source now or gtfo

>>58153096
Not if it's half the price. But I don't see how they'd be selling an 8core for <$150.

cant wait to see the tears from amdrones on release day

Seems pretty competitive.

Even better when you remember AMD typically allows overclocking, provides all new instruction set extensions on all of their processors, and also allows virtualization features. Plus they're generally cheaper than Intel processors. Another nice thing is the AM4 socket works for all tiers of processors so you can buy a cheap one and upgrade to a better one when you have more money.

Discussed here dumbfucks
>>58138937

>>58153096
>>58153336
>>58153331
>>58153300
Look at the clock you fucking retards

File: 1475453515129.jpg (45KB, 500x431px) Image search: [Google]

45KB, 500x431px

>>58153096
looks like AMD lied again.

Hope they go bankrupt this time.

File: 1472439659580.jpg (200KB, 837x892px) Image search: [Google]

200KB, 837x892px

>w-wait until we overclock them

>3.15ghz
>overclock past 4GHz
Communist poorfag pajeet scum BTFO.

>>58153342
Nice defense.

I'm posting from a desktop with an Intel CPU and nVidia GPU. I'm not AMD fanboy you dumb faggot.

I want to know why these images look so fake. Why is the graph so shit when the text is so clear.

It screams fake and you'd have to literally mentally retarded to think otherwise.

SOURCE
O
U
R
C
E

>>58153355
Games cannot into more cores so it's not really surprising. The games use what they can and the rest of them go idle and contribute jackshit to the score. They do however affect the turbo because they still draw power and give off heat.

>AMD promises some new mystery product that supposedly will BTFO the world
>Gives no information for months aside from cherry picked benchmarks
>it turns out to be complete and utter bullshit
>AMD shills on /g/ make up excuses to buy it anyway "it'll be good in the future!", "do you wanna support a monopoly/da joos?"

oh ah boy who would have expected such an outcome? It's almost as if this isnt what they've been doing since bulldozer.

>>58153382
>It seems fake
>So it is
Just figuring out that the chines are shit at making graphs?

>>58153419
>even the leaked benchmarks show it delivers on everything promised
>"hurr i guess i'll just convince myself it's bad"

brain addled consumerist fanboy cunt

>>58153377
That is it's stock clock
All of them are at stock clock, you fucking retard

The chart seems to suggest that Zen is almost identical if not better than Intel's offerings clock for clock.

Zen may actually exceed my expectations holy shit.

File: 572457635.jpg (196KB, 900x509px) Image search: [Google]

196KB, 900x509px

>>58153478
>Delivers everything promised

>>58153490
Are you fucking retarded?
Holy shit you're bad at math

>>58153490
that's assuming it can clock significantly higher than 3.4ghz

I need a new CPU and probably won't be able to afford Zen (since I also need a new mobo and fan; my Phenom II X4 isn't carrying me well anymore.

Is the 8350 a good budget buy?

>>58153096
No 7700k. :^)

>>58153490
In order for it to reach intels clock for clock it would have to reach above 4.4Ghz
Like >>58153513 said, you suck at math

>>58153500
I won't buy anything with retarded gaymur name like ryzen.

>>58153523
just wait. AM3+ is dead and piledriver sucks. Intel's low end CPUs are dog shit also.

https://twitter.com/FPiednoel/status/812496510906560512

AYYMD IS FINISHED & BANKRUPT

>>58153527
The 7700k is barely faster than the 6700k and kabylake fags can't accept that they waited for nothing, and coffeelake is mobile only so more waiting

>>58153527
seeing as the 6700K is already there, that would be redundant

File: 2016-12-12-image-21.jpg (107KB, 1442x753px) Image search: [Google]

107KB, 1442x753px

>>58153529
AMD specifically stated that it goes above 3.4GHz. This chip is 3.4GHz in turbo. Either its not the top or AMD lied.

File: amd2.jpg (52KB, 568x612px) Image search: [Google]

52KB, 568x612px

>>58153542
fucking dumb frog

>>58153560
Read harder anon
I said 4.4 GHz not 3.4
in order to read above 118% it needs to hit 4.4GHz

>>58153096

>on par with the 6600
>lower power and lower cockspeed (kek)
> will demolish Intel cock for cock.

The hype is dead. Now we truly know that Zen is great product. Further hype isn't needed.

>>58153620
>on par with the 6600
>will demolish clock for clock
on what fucking planet

>>58153327

Actually its the other way around. Here you have a product comparing with high end Intel and costing half the price.

And its running cooler and with lower clocks. Meaning clock for clock it has higher IPC then all of Intel.

If it clocks to 4.2ghz it will demolish everything.

I'm fucking stoked. Now if Vega isn't a let down like the 480 was. We are good to go.

>>58153628
On par with a chip that is higher clocked and has fewer cores in games which demand higher clocks and fewer cores. Looks good to me.

>>58153644
in what way was the 480 a letdown

it takes a big steaming shit all over the 1060

>>58153483

Dumbass. It's clocked slower meaning it will rape them clock for clock.

>>58153662
No it wont
to hit the performance of the 6700k at 4.0ghz the amd chip will have to be at 4.2ghz

do the math you autist

implying it can even overclock to 4.2

>>58153590

Nope. It needs 3.65668 to compete.

>>58153661

I have a 290x so I had no reason to upgrade. For me it was a let down since I wanted 1080 performance.

>>58153676
> implying it can even overclock to 4.2
It would, and you don't even need to pay goytax to OC ayymd chips.

>>58153678
You went from it's minimum 3.15 not it's max 3.4 anon
Do the math again

>>58153698
out of a gpu that's less than half the cost of a 1080

the real disapointment is that vega is taking fucking forever to come out

>>58153700
>It would overclock by .8ghz
>AMD processor
>Adding .8ghz
Don't make me laugh anon

>>58153728
We don't know shit about Vega. That's the real disapointment.

Where did all these pessimistic anons come from? We barely know anything about it much less how it overclocks and anons are already saying it overclocks like shit.

>>58153750
That should tell you everything you need to know
We don't know a single fucking thing about this product from a company with a track record of massive disappointment, yet people are still building hype

I'll stand back mildly interested

>>58153760
/g/ is rightful Intel clay.

>>58153760
Because AMD has a history of shit overclocking besides the 95whateverthefuck that overclocks to 5.ghz but gives you shit performance

>>58153766
>with a track record of massive disappointment
The only actual disappointment was faildozer.

>>58153760
they're children who don't remember P4 and Fermi

they started paying attention to the industry halfway through bulldozer's lifespan so to them intel and nvidia are god and amd can't do anything right ever

>>58153246
samples been around since May, who the heck knows form when that one is
judging by clock looks like summer, rumors of that clock were exactly at that time

>>58153789
And polaris
And bobcat
And jaguar
And their APUs (not all of them)
And the Pro duo
And the Fury x
etc

>>58153178

Not him but why not? It's faster, it's cheaper, it's more efficient. I would for virtualization needs.

>>58153805
>poolaris was disappointing
Why? I'll give you Fury X.

>>58153805
Bobcat and Jaguar weren't bad though. APUs sell well and they're extremely cost-effective which is why AMD likes them so much. Depending on who you ask the other stuff might be true.

>>58153819
480s killing mobos and not giving the performance we expected from the keynote
and why would anyone buy any other graphics card than that

What about this self overclocking AMD was talking about, you know the one who learns and changes based on what it's asked to do and how good your cooling is... seems like something that might be useful in you know ... games?

I kinda doubt that was turned on in these benchmarks

>>58153801
>P4
Oh, the times Intel went on literal suicide watch and bribed OEMs to stay afloat.

>>58153734
Is this b8
Do I have to remind you of the 8350?

>>58153805
>Polaris
It's a cheap card, what do you expect?
>Pro duo
They said they would give us two furies duct taped together and we got two furies duct taped together.

>>58153848
>killing mobos
XD

>>58153819
tech press blew it out of proportion
reference cards had shit coolers

that's enough to cement a fanboy's opinions forever

just trigger them with this and move on

http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/73945-gtx-1060-vs-rx-480-updated-review.html

>>58153854
and sabotaged their C/C++ compiler

>>58153856
8350 is still shit anon, have fun with your bottleneck
>>58153862
Expected better than what they gave us
The pro duo has shit frametimes
>>58153862
Yes it's killed mobos you can't just deny it because you don't want to believe it

>>58153870
> tech press blew it out of proportion
> reference cards had shit coolers
That happens with every card ever. Period.

>>58153715

desu i just wrote in a random number, was in the toilet and cba to do it lol.

at 3494mhz its comparable to the 6600k (which btw is running at 3900mhz)
at 4130mhz it should be at the 6700k level which is running at 4.2ghz.

Now the 6700k costs 350$
And we are assuming from the model number that this is infact not the black edition meaning you get a 8 core for the same price as the 6700k.
And lets take into account AMD mobos are cheaper than intel, and with rebates etc, the cpu might drop 20$

you basicly get a 8core cpu for the price of a 6700k. and if it clocks to 4.2ghz, you get good enough performance.

Also to add, DX12 and async compute which will be more cpu dependant. we might see more and more games run smoother with moar coars.

Now lets break down this shitty chart. on what planet does the 6700k perform 20% faster than the 6600k in games lol. that just doesnt make any sense.

So before we stroke our dicks in pleasure, lets wait for official reviews. then decide what to buy? and hopefully the fucking trolls wont be en mass by then.

>>58153876
>The pro duo has shit frametimes
how to spot paid shill

>>58153900
How is criticizing frametimes being a paid shill?
Playing 60fps with frametimes all over the place is not smooth, not to mention the pro duo wasn't supported in half of the adobe creative suite

>>58153766
>track record of disappointment
>64 bit extensions for x86
>removing the inefficient FSB
>pioneering IPC performance while Intel pushed "muh clockspeed"

AMD has a long history of innovation. Even Bulldozer wasn't a complete failure, since even now it can compete with Intels current generation in certain metrics.

Certainly AMD hasn't had good overall performance parts for quite a long time, but these parts are all a result of misreading the market, not technical incompetence. Saying that Bulldozer indicates what Zen will be only demonstrates a complete misunderstanding of processor technology.

I am not saying Zen will outperform Intel's parts across the board, but we have every indication they will be competitive.

>>58153377
Stop it, its just pathetic at this point.

>>58153096
Fake and gay

>>58153728

but the 290x is faster than the 480 or in some games equal.

again no need to upgrade unless i get a fury X or nvidia, and im never getting nvshit.

>>58153850

its a pre-release cpu, not even finalized yet.

the retail one will probably be a tad faster.

>>58153916
Just because they have innovated doesn't mean we ignore their shortcomings

>>58153948
>we ignore their shortcomings
And no one does.

>>58153947
>Retail will be a tad faster
when has this ever happened
retail has always been slower than the binned chips for testing

>>58153967

its an engineering sample...

>>58153911
>Playing 60fps
it wasn't made for that. are you going to repeat all meme you ever heard?
also not a single motherboard was broken, even after sever tests nobody could replicate burning even cheapest boards

>>58153967
are you for real?
ES and binned chip is two greatly different things

>>58153979
Dude engineering samples are always faster than the finished product
Do you work with any tech company because you would know that. My company's engineering sample wafers are always better
>>58153985
>not a single motherboard was broken
https://community.amd.com/thread/202410
>inb4 doesn't count

>>58153993
>Engineering sample and binned chip being greatly different
>A month from release
no

>>58153948
Bulldozer was the only disappointment I've seen. The Fury line performs as it should for the price, as does Polaris. There have been some hiccups, of course, but nothing that hasn't been done by AMD's competitors. Intel released Netburst, but as we can see, that didn't prevent them from fixing their mistake in the next iteration. Nvidia has released hot inefficient cards, have released multiple card killing drivers, and have blatantly lied about their cards in the past, yet nobody assumes these things about upcoming Nvidia parts.

AMD hate is a meme. It's boring, really.

>>58153660
This, anyone who doesn't see that AMD is now on par with intel is utterly retarded. This chip is clocked to 3.4ghz while intel ones are clocked higher, at the same clockspeed they will perform the same

Really looking forward to Zen now. I really hope it manages to exceed my expectations like these charts indicate.

>>58153096
Well, what the fuck did you expect, it's only a 40% increase in IPC

File: cover.jpg (609KB, 1280x1280px) Image search: [Google]

609KB, 1280x1280px

>>58153785
The 8320 can overclock quite a bit but it doesn't really make the chip relevant outside of very specific situations. I know radeon is totally spereare from their cpu with even a $60 AIO slapped on a rx 480 it can pick up over 15% pref which is nice compared to the 10% on Nvidia unless you lower the resistance on the filters. It's pointless to speculate, we haven't seen their new overclock tech yet so let's wait. Unless you have stocks but I'd be selling at this point to be safe.

>>58154081
>>58154053
>Being this delusional
It's ok, you'll get the idea when it releases and then you'll be the same old idiot when the next processor releases

>>58154081
I don't think it will be revolutionary, but I do believe pricing will become sane again. Ten years ago I built a midrange system on a single paycheck while I worked part time in high school. You can't do that now, but hopefully by the summer the competition will be on.

>>58153527
>100℃

since you really don't need more then a 6600k for gaming if It's on par with that and cost the same as a 6700k but with 8 cores and 6900 level performance on tasks that actually require a lot of CPU horsepower... how is this 'ZEN DOA!'

That's even assuming all of the ryzen dynamic overclocking was pure lies and won't effect games at all

>gayms

>>58153527

just use the 6700k as reference there bud. just add 1% increase in performance and add 40c to the temperature ;)

>>58154173
>add 40c to the temperature ;)
Is meme lake THAT bad?

>>58154219
For once skylake fags were right and kabylake fags were wrong
It's 1% faster than the 6700k and overheats like a motherfucker

>>58154219

its not that bad, unless you OC it to 5.1ghz, then its that bad.

but its basicly a 6700k with higher power consumption and shitty metal plate ontop so it runs hot.

delidding it actually made it lose 30 degrees lol. which shows the true power of memelake cpus.

but delidding isnt exactly safe

What's with the shitposting in this thread? It's an ES with way lower clocks, of course it's not gonna compete with something clocked way higher in something like games, even Intel's $1500 jewgrab doesn't do all that well.

Will there be dual core Zen chips? Dual core chips could lead to a lot of product tiers:

2 cores
2 cores + SMT
4 cores
4 cores + SMT
8 cores
8 cores + SMT

>>58153096
I am still patiently waiting.

...and still cannot believe that only two companies in the whole world are competing in such important technological advancement

>>58153529
6900k all cores run at 3,5 ghz (turbo enabled) with 3,7ghz max boost (single core). These benchmarks compare Zen running at an unknown frequency, it might run only at 3.2 ghz. And we don't know what other performance issues it might have. These benchmarks show that AMD really has IPC near Broadwell (5% margin). If there will be a Zen chip with stock clock over 3.4ghz and tdp 95w it would be a no-brainer choice over the 6900k.

>>58154278
Sure thing bud :^)
AMD will definitely deliver this time right

>>58154258
I don't think AMD will segment the market that much. Or at least I'm hoping they won't. I'd really like to see SMT enabled across the board.

>>58154246
Intel greed knows no bounds. Why the fuck did they stop soldering their fucking chips since Ivy.

>>58154278
We already know there will be. Lisa Su showed them off last week.

>>58153096
LOL

What good is IPC when ZENSHIT can't overclock more than 3.1GHZ LOOOL

>>58154292
SMT can cause a slight loss in performance in very specific cases and it does causes a slightly higher than normal power usage and higher temps (but more efficient utilization of resources). I think AMD will allow people to disable SMT like Intel does but they might roll SMT out across the board.

>>58153760
To be fair it's been a long, LONG time since AMD had a CPU which can compete with Intel properly. On the graphics card side I feel most of their latest products kind of underdelivered, Fury X looked like a monster but didn't perform like one, nor did it OC well, Pro Duo is so expensive they had to spin it as a "semi-pro" card, but CF support has been quite lacking lately (as in the last 2 years), RX 480 came out with 2013-tier performance, not much OC headroom and isn't even amazingly efficient either. It's not that hard to see why people in general wouldn't have amazing trust in AMD hardware at this point. Even though Radeon and AMD CPUs don't actually have much in common, they both affect trust in AMD's brand.

Zen OC is obviously unknown, if it ends up at 3.4GHz base + higher boost for 8C/16T in a 95W envelope, that could be quite nice considering Intel has 140W TDP for CPUs in the same ballpark clock wise. Going by the IPC it seems to have it would need to compete with Intel in OC as well, so basically do 4.5+ fairly reliably, maybe 4.6-4.7+ with more effective cooling like a custom water loop.

>>58153096
>8c/16t gaming benchmarks are out
>we are going to pair it with 4c/8t CPUs
>we are going to expect them to be on par

this is how multicore CPUs scale with non-scaling programs

>>58153096
>everything above is clocked much higher

Are you an idiot?

Is this the thread where OP pretends to know how ES are made and for what reason?

>>58154419
Yes.

>>58154387
worse than that the skylake i7 is running at a 3.7ghz boost clock and the Zen chip is only boosting to 3.4ghz. So A skylake i7 is only 10% faster then it has a slightly less than 10% higher clockspeed.

This means we have at least for vidya gaming purposes a zen chip that has acheived IPC parity with skylake or is so close behind it as to not matter.

so imagine a zen 4 core hyperthreaded and unlocked chip that can overclock well into the 4.2-4.4ghz range if not up to the 4.9-5.0ghz of a watercooled skylake/kabylake but unlike the currently 300+$ i7 "K" parts you get this equivalent zen part for 199$? or maybe they go completely insane and make it 159$?

AMD is already offering cheaper motherboards that will allow overclocking of the CPU so you wont have to fork over extra shekels for a special "Z" motherboard just to overclocked your overpriced "k" chip.

Imagine a 70$ low end single PCIe mobo that can overclock a decent 4-core i7 equivalent Zen chip thats only 150-200$......

Thats a full base system that can drive framerates at 1080p/1440p/4K at 60hz/120hz and even 144hz with the right GPU for literally 220$-270$

To get the same perf from intel you need a 120$ "z" motherboard with a 300$ "K" i7 chip

now add the cost of an X-series chipset and a 6-core or 8-core 'broadwell processor?

for AMD its the same cheap a dick 70$ mobo or 100$ for a higher end one with multiple PCIe slots for crossfire/SLI and then 299$ for a 6-core and 499$ for an 8-core all with hyperthreading and unlocked for overclocking by fucking default.

Even if you love sucking nvidias green cock you would buy AMD just so you can take the money saved and put it into a more expensive nvidia graphics card

>>58154419
No
You foking retar

File: 6436536424.jpg (18KB, 394x394px) Image search: [Google]

18KB, 394x394px

>>58154392
>XD we should just underclock all the others then AMD wins go team

>>58154577
First of all you don't know what boost is, 3.4 boost is for one or two cores, not all of them, 4 cores don't boost or boost very slightly (like 100MHz)

>>58154605
>what is ES
(you).

>>58154577
fix your fucking format
it's painful to read.

>>58154619
>Just using buzzwords to make yourself feel right
Kill yourself
Engineering samples don't matter as much as you'd like them to newfag

>>58154642
Okay, have another one. (you).

Intelfags must have been hit really hard when they're trying to pretend low clock ES are the same as chips that are gonna ship.

>>58154664
Nice argument :^)

>>58154611
in that case its even worse for intel because the IPC gap is more extreme at least for multi-core

these are not the final clocks anyway its a test sample and the final chip will bare minimum be a 3.4 base clock with a boost likely in the 3.7-3.8 range with OC potential at least to 4.3-4.4ghz

I dont expect the 4.7+ghz of skylake K chips but it will be close enough behind that you would buy this over a shitty i5 or a lower binned and locked i7 and the only people who would pay the premium for a 4-core i7k would be neckbeards who need 143fps instead of 139fps

>>58154689
Stop posting the same shit about engineering samples you retard
Those are the clocks of the finished product, that's what it's going to be released as.
Engineering samples aren't going to be different from retail a month away from release you fucking idiot

Show me proof that this is an underclocked engineering sample as you say it is

>>58154642
It's an engineering sample, which means it's not final silicon. We have no idea when this sample was produced. That we're seeing it now is irrelevant. We've been seeing ES at 3.15 ghz for weeks, if not months. AMD has stated that Zen will clock at a minimum of 3.4ghz before boosting. That came from the CEO. AMD would have been more likely to declare bankruptcy than send out their CEO to tell such a stupid lie.

>>58154728
>3.1GHz base
>3.4GHz boost
It's in the fucking article

>>58154728
>Engineering samples aren't going to be different from retail a month away from release you fucking idiot

AMD doesn't necessarily have to give reviewers a sample chip from the most recent batch. This might be a chip from before the final clocks were decided.

>>58154770
>>58154780
they only confirmed that ONE chip will have a base clock of 3.4 not all having the same base clock

>>58154728
Finished product? You got fucking benches of 2.0GHz ES floating around, that's what they are, ES, they can be a year old because they're sent to OEMs for testing and debugging way before release, and this one is a notably lower clocked one from what's supposed to release.

File: 1481934400706.jpg (61KB, 394x473px) Image search: [Google]

61KB, 394x473px

BULLDOZER 2.0

>>58154802
Well these clearly aren't from those batches :^)

Watch when this shit releases you'll still be claiming engineering sample fucking lmao

>>58154801
yea, the 8c/16t

>>58154821
Are you just pretending to be retarded? It's almost like you believe there aren't Skylake ES out there with sub 3.0 clocks or something.

JUST

>>58154801
Chances are if your 8 core has 3.4 base clock your 6 and 4 core will have much higher base clock under the same TDP following basic thermodynamics.

>>58154830
And that chip is limited to TDP

a 6-core and 4-core model will have more wiggle room in finalized clocks for their TDP envelopes which means you can expect an extra .1-.2ghz or so on their base and boost

>>58153096
kek we already know that its an ES chip running on 3.15ghz
op dont even try to see if there is other thread running with the same thing..

>>58154857
yea, and base speed of 8c/16t is 3.4ghz

>>58154839
There are
But none of them were used for benchmarking, ever

File: brutal_child_fisting.webm (2MB, 718x412px) Image search: [Google]

2MB, 718x412px

>>58153096
>16 zen threads = 4 thread from consumer i5

>>58154870
There are two chips with 8c/16t

>>58154885
>gaymen

>>58154889
yes, and both of them have 3.4Ghz base speed.

>>58153096
Can you do math? The 8c/16t Intel is 10% faster also running 10% higher clocks.
Also lets assume this is a best case scenario ES with a fully working mobo without debugging hooks, up to date microcode, thread scheduler OS support and fully enabled memory channels/cache, which is possibly a pipe dream.
What does that tell you about the IPC?

>>58154885
Gaming performance, the majority of the threads don't contribute anything at all or just negatively affect the score.

>>58154906
No they don't
The second one is cheaper and has a base of 3.2 to 3.4

>>58154870
which is completely decent. Even if the boost is only 3.6ghz a 3.4/3.6ghz 8 core chip with SMT and an unlocked multiplier will sell very well if its as cheap as we are hoping. Assume a lower binned basic 499$ version of the 8-core with hit and miss OC and then maybe 599$ for a "black" edition with higher binned chips certified to reach certain clocks at specific voltages.

>>58154912
>Source = My ass

>>58154924
Welcome to /g/.

>>58154924
Wow funny joke XD upboat
4chan jokes are funny XD kek newfag underage XD

>>58154912
All zen chips have a guaranteed base clock of 3.4 it was announced by Lisa at Horizon

>>58154876
so? who the fuck cares lol
truth is the magazine states that is an ES chip with a core speed of 3.15ghz not even close to the 3.4 lisa said
so get your shill shit and gtfo

Was anyone honestly expecting it to beat Intels top processors? It's the best AMD has ever made but of course it's not gonna measure up to Intel. Wtf were you expecting?

>>58154945
>>58154924
Don't samefag

>>58153096
Source of the images. http://www.guru3d.com/news-story/french-magazing-posts-engineering-sample-amd-ryzen-processor-benchmarks.html

>>58154958
>Guaranteed
When have they ever let us down right anon
AHEM bulldozer COUGH COUGH

>>58154912
We only known base clock and ONLY base clock of one SKU so far.
Also when talking about turbo nobody even mentions that AMD and Intel do turbo differently just like they count TDP differently.
Everyone just assumes because it says 3.4 Turbo in the OP that all cores turbo to 3.4, they don't, only one or two do.

>>58154986
a year before bulldozer was out the former ceo warned that the ipc will be horrible
so only stupid people still complains that amd lied and bla bla..

>>58153096
benchmarks are made for intel cpus

>>58154986
They know they can reach a sustained multi-core of 3.4 and lie bout that live at their own press conference to their own SHAREHOLDERS would open them up to massive lawsuits.

you are retarded and cannot into Stockmarket/share prices

>>58154965
I'm not expecting them to beat Intel but if this chart is legit these chips seem to have IPC that is very close to Intel's and could even be superior. I'm going to wait for retail batches to come out and for benchmarks to verify that before I get too excited though.

File: itwasallme.jpg (149KB, 398x1031px) Image search: [Google]

149KB, 398x1031px

>>58154966
Jokes on you, it was all me

>>58155002
>O-Only stupid people say AMD lied
>P-Previous CEO admitted it
>A-AMD would never lie
>T-T-T-T-T-T-T-T-T-T-T-Team Red

>>58154246
>but delidding isnt exactly safe

Ok... So we know from the marketing that AMD CPU will "auto overclock" based on the cooling solution.

Does AMD cheap out on the paste compound of their CPU making the temps go up ?

Has anyone delided an AMD CPU before ?

>>58154988
Lisa Su said the minimum clock for any Ryzen chip will be 3.4. But I guess you'd know better than AMD's CEO, right?

(also, since we know there will be two 8c16t SKUs, that means the higher binned one will be clocked at 3.5+)

>>58154988

Turbo varies wildly on intel. for single threaded tasks you get 1-2 cores at the top end of the turbo but for 4-cores maybe half of that increase on sustained or barely any increase at all and just baseline for multicore

on a 3.2-3.7 intel base/boost you should expect 1-2 cores at the 3.7 and all cores at 3.4-3.5 if your lucky

>>58155050

AM3+ is soldered.

Hey, wouldn't the performance be the fucking same as a 6900k assuming equal clocks?

What are you people complaining about?

>>58155076
But Zen die is half the size.

>>58155076
>AM3+ is soldered.
Care to explain a little better anon ?
Like if i were retarded.

Is it good ? is it bad ?

>>58154988
differently...
even the ES chip on ops shitty bench is showing a tdp of 93watts...
kek
also the fuck are you saying? the magazine states clearly 3.15ghz base clock 3.3 turbo all cores 3.5 turbo single core..

also 3.4 is the base clock as per lisa on the final product so..

>>58155110
It IS good. Remember Sandy Bridge?

>>58155093
No because Zen can't go higher than 3.4GHz at all because GloFo is shit

File: ABG2.png (801KB, 800x900px) Image search: [Google]

801KB, 800x900px

Literally means nothing as long as we don't know the actual fucking price.

If it's gonna cost $150 then it's the greatest product since 2500k and Athlon.

If it costs ~$200+/- then it's gonna be a nice alternative to 6600k but it's probably gonna lose same way 480RX lost to Pascal cause of brand recognition.

If it costs $300+ then Nvidia/Intel fans were right once again and Pajeets were lying as always while /g/ is gonna keep being biased towards that shitty company.

I honestly want them to be good, not too good because I don't really trust AMD enough to buy their products since 2010 but I want them to be good enough so I can see price drops on 6600k/6700k.

>>58155110
He's saying the die was soldered to the IHS. If you tried to remove the IHS it would tear the die off. you don't need to delid processors that are soldered though.

>>58155110
Well, it's better for socket to be soldered on, otherwise it wont connect properly to the motherboard properly.

>>58155121
Whoa, 3.5GHz single core? That pretty cool, adaptive clocking seems to work nicely if it's working on this ES.

>>58155093
It is the same, AMD demonstrated that last week. They're upset because some of them are trolls, some of them are idiots, and some of them are likely to actually be paid shills.

>>58155127
lisa said in front of the shareholders zen will have base clock of 3.4ghz
anon on 4chan who obviously knows better
>zen wont clock higher
literally the embodiment of stupidity is reaching critical levels around here

>>58154096
>Source: my AMD ravaged anus.

480RX can't gain 5% without crashing on 99% of customers.

>>58155156
Stop it with the engineering sample anon you've been samefagging this entire thread

>>58155126
>>58155138
>>58155146

So soldered to the IHS allows for better heat flow, therefore better the cooler can extract more heat & faster.

Thanks anons

Why are Intel fans foaming at the mouth pretending Su didn't say Zen will have much higher clocks like 2 weeks ago?

>>58155169
>source?
yeah no as always coming out of your ass

>>58155169
>Source: 99% of the facts etc etc etc

>>58155184
That means OC potential which is why Intel stopped soldering their chips since Ivy.

>>58155174
You seem to be praying real hard for these to be the final clocks, don't you?
Your praying seems to be pointless though.

>>58155203

Aren't the enthusiast chips still soldered? Xeons most certainly are soldered (but then again Intel nuked overclocking xeons).

>>58155184
The only reason AMD solders the IHS is because they can't afford to RMA de-lidded CPUS

>>58155131
Their GPU's are better to be blunt. I have a gtx 1080 and an RX480 and I vastly prefer the Crimson Drivers and user experience of the 480. Nvidia Drivers make me wanna tear my hair out and if Vega is anywhere close to the 1080 in speed then i will probably sell it and switch.

As for CPU's theres not much for AMD to fuck up especially after all the hard lessons they learned from the bull dozer arch. Just get the single-thread and IPC close to intel and give me automatic SMT and unlocked multipliers and give me more open-ended options for motherboards on your AM4 socket and make it cheap enough to be worth my while and i will switch no problem. What else does intel have? specialized instruction sets that normal users would never fucking use?

>>58155188
No, literally only 3.4ghz
>>58155220
You seem to be praying these are engineering samples

>>58155131
This is the highest configuration Zen. It will be one of the higher priced chips but probably not $1000+ like how Intel fucking prices their stuff.

4 cores are probably going to be the introductory tier at <$200, and then 6 cores above that, and then 8 cores above that. They might make dual cores at budget tier prices too.

>>58155221
>Aren't the enthusiast chips still soldered?
Yes but average consumer gives 0 shits about enthusiast chips.

>>58155232
>I vastly prefer the Crimson Drivers

Stopped reading it there. At least make it subtle.

>>58153948
> we ignore their shortcomings
It's not the same ignoring shortcomings than putting all pros and cons on a balance, things are not always black or white. Just saying.

>>58155244
>You seem to be praying these are engineering samples
Maybe because they are?

>>58155221
I am hearing talk of a scandal involving Kaby lake where they de-lidded the i7 and it caused a 30C temp drop which means the thermals of the package itself are all kinds of fucked up. add onto that the Zero IPC gains over skylake and its a fucking bust.

>>58155244
>only 3.4 base
>from 3.1 base
>on a 8 core
That's a 10% MINIMUM frequency increase per core you stupid fuck.
Even if only 4 cores are used that's still a considerable increase when your competition is withing spitting distance.

And what the hell do you think these are but ES? Retail units? Are you fucking retarded?

>>58154096
You can tell that AMD shills are getting absolutely desperate when they don't even try to be subtle with their fucking shilling. Btw buddy, not even gamernexus, the most obvious AMD website out there could hit 1400 core on the hybrid cooling.

>>58155294

That wasan't about the TIM itself just Intel ejaculating a fucking mountain of adhesive to keep the heatspreader in place and naturally glue doesn't have good thermal properties for this sort of thing.

>>58154923
>certified to reach certain clocks at specific voltages
What is the point of such a CPU? Why not sell it at the higher clocks at stock, like the 9590 or w/e it's called?

Looks guys, a 2.2GHz skylake Intel is finished these are retail units not ES http://chinese.vr-zone.com/131337/intel-skylake-s-cpu-sampling-greenlow-platform-will-on-schedule-as-2015-q2-10192014/

>>58155304

>gamersnexus
>a site that slaps a 9590 on a 970 mobo not rated for it and then wonders why the 9590 posts worse performance than an 8370

>>58155304
lol so you base your facts out of something that happened on the launch and there wasnt a 3rd cooling solution yet?

come on dont be so fucking obvious stupid 1400 on really good air cooling is izi pizi let alone on water

>>58155255
Not a Troll. Crimson was easy and problem free. All i ever got was silly problems like a corrupted mouse icon or my compass getting wonky in fallout 4. Nvidia GeForce Drivers literallty Skull-Fucked my entire operating system and DDU doesnt even get rid of all the traces of their cancer.

Crimson IS vastly superior to GeForce. Nvidia might have the better hardware right now but AMD has the better software and that is no longer in dispute especially for me when I have having less trouble and a better experience on a fucking 240$ RX 480 than I am on my goddamn 700$ gtx 1080. That should never fucking happen.

>>58155317
probably to win enthusiasts and overclocking over?

Generally each company makes at least 5-10 ES models before the product hits the shelves and distributes it to OEMs for testing, this is just one of those, we already had leaks of differently clocked ES months ago.
I wouldn't hold these benches in high regard because who knows how old the sample is and what and whatnot is disabled.

>>58155344
You win enthusiasts over by selling a more expensive CPU with the same stock clock as the cheaper version, but a "guaranteed" OC? What exactly is the difference between a guaranteed "OC" and the stock clock? Both of them are the clock the manufacturer guarantees the CPU will run at.

>>58155416
In theory, you can OC your CPU and boost will be at the same range.

>>58155323
Those clocks would be acceptable for a 16+ core.

>>58155341
Yes because it never happened. You're not even being funny, it's just sad shitposting.

>That's no longer in dispute.

Is this some sort of a reversed day? Are AMD fans on /g/ truly this delusional?

>>58155490
Not that anon, but you'd have to be brain dead to think Wattman isnt amazing.

Granted, the name is dumb.

>>58155477
>boost will be at the same range
I don't understand what you mean

>>58155531
This state of mind seems to be common in this thread, yet people still talk shit about stuff they're ignorant about.

>>58155531
let's say, boost is 1ghz, you can OC your base clock by .5ghz and still have 1ghz boost.

but shit is not very well explained just now.

>>58155490
>butthurt nvidia fanboy detected
>gave all his allowance money to jewvidia
>to get the best performance on his 900$ 144hz G-sync monitor for battlefield-whatever-version-it-is-now
>still uses 90% of his PC time to jack off to chinese cartoons and sperg on 4chan

>>58153096
>$1000 Intel doing worse than a $300 Intel

Wow

>>58155304
Here you go >>58155336

>>58155598
It's all in the clocks, vidya games barely scale above 4 cores.

>>58153096
>benchmarking CPU in GPU-heavy applications
Call me when prime95's are out.

>>58155615
It doesn't matter when Zen is confirmed for 3.4GHz MAX even if its per core perofmrance is higher it can't match a 4.6GHz intel core

File: 34D7F26E-14C6-4ABE-ACFF-835F9D71A796-704-000000F7CC4FA4B1_tmp.jpg (110KB, 590x306px) Image search: [Google]

110KB, 590x306px

>>58155283
>>58155297
>T-They are engineering samples
>A-Amd will win you'll all see
Screencapping this thread to run in your faces when AMD lets you down


Again

>>58155520
No. I don't. The idea of even having to touch your voltage in 2016 just to gain any significant performance boost is outdated and is only another point on the never ending list of AMD issues. Nvidia OCing might be simplistic to the point where even a drooling retard could do it without ever being in a real risk of fucking up his GPU but it's literally a selling point. You buy the card, install Nvidia Inspector or some other software with gaymen UI like MSI, increase the power limit, core and memory clocks and get free 10% out of the box without ever touching voltage. Meanwhile 480RX owners, especially the ones who bought reference cooled cards sometimes had to underclock their cards in order for them to even work within sane noise and temp loads. Getting any sensible results in OC means touching voltage of 480RX. You really don't sell it to me by saying that AMD allows you to risk your warranty in order to stay somewhat competitive against Nvidia's 2 years old hardware.

>>58154811
I wish Bulldozer had performance parity like this. AMD would have been filthy rich probably.

File: FD468783-E445-4BBE-ABD4-27216C8D3162-2376-000001C22A6939CF_tmp.jpg (24KB, 500x478px) Image search: [Google]

24KB, 500x478px

>>58155757
>performance parody like this
>AMDumbfucks are this delusional

>>58155775
>>58155701
Great filename, says a lot actually.

>>58155703
It's still silicon, of course you have to touch the fucking voltage. If you raise clocks without voltage you're only tapping into what the manufacturer reserved for stability.

>>58155787
That's a phone file name
The first one is android the second is and ibone

>>58155828
Other way around anon first is iPhone second is android

>>58155563
I see what you mean, but I doubt it will be like that. In fact it would be pretty horrible if you didn't have complete control of the CPU when you OC. It will most likely work like it does with unlocked Intel CPUs where when you OC you basically set the maximum boost clock and either always keep the CPU running at max clock all the time or only allow it to go below that, not above. That's the only way it makes sense once you take the silicon lottery into account.

File: .png (87KB, 299x288px) Image search: [Google]

87KB, 299x288px

>>58155828
>>58155849
Android preserves original filenames. And neither of these look like iPhone names(it's either image.jpg or DSC_*).

>>58155695
Intel doesn't have 4.6GHz stock cores. You'd have to compare it to OC'd Zen, which we actually know absolutely nothing about at this point.

>>58155703
Maxwell and newer can't undervolt, I mean it's great for idiots but fine control is still preferable.

>>58155695

>Zen is confirmed for 3.4GHz MAX

Are you retarded? Don't answer its a rhetorical question - you clearly are retarded.

>>58153096
All I see is AMD shitwrecking intel.

>>58156001
Delusion

>>58153355
The 6900k has 8 cores too but still loses against the 6700k. This is a single core benchmark.

>>58156001
Their 8c/16t cpu will cost at least $500. At least according to http://www.pcadvisor.co.uk/new-product/pc-components/amd-zen-processor-release-date-price-specs-features-3643552/

Which when it comes to my britbong prices was pretty much always on point.

So a $500 cpu is losing to a stock $220 Intel cpu in actual performance that interests 90% of buyers. Your post really fired up my synapses anon.

>>58156057

As opposed to Intel's own 1000 dorra chip losing to a 200 dorra one?

>>58156084
In gaming?
Yes
You fucking moron

>>58153529
>compare intel CPU that is clocked 17% higher to AMD cpu
>intel CPU performs 20% better
really makes you think

>>58156136
Doesn't matter when AMD's hot garbage can't reach those clocks :))

>>58156136
>implying AMD can get up that high in cocks

fucking been saying this all along,

all those cores are useless for general everyday shit and gaming.

was it any fucking surprise that AMD would release stuff that highlights Zen's strong points like rendering shit ?

>>58156155
AMD hasn't released a single CPU line that couldn't overclock, even their short-pipeline designs could net a decent clock.
Stop being retarded and sit down.

>>58156173

Well according to /g/ gaming is the only purpose for a cpu to exist.

File: amdvsintelsteamsurvey.png (12KB, 1153x61px) Image search: [Google]

12KB, 1153x61px

>>58156084
Well done anon. You proved me that you shouldn't buy a LGA2011 socket Intel CPU for applications that are optimized to use 4 cores and see best performance gains in higher clock speeds, not amount of cores. You're this close to discovering America when it comes to casual/gamer market and their choice of brand.

File: Ryzen-Magazine-Cover-Page-872x1140.jpg (195KB, 872x1140px) Image search: [Google]

195KB, 872x1140px

>>58153318
>>58153357
>>58153382
>fake
0/10

http://www.cpchardware.com/

http://www.comptoir-hardware.com/

I remember that fool at anandtech forums and RWT claiming Zen 8 cores won't even reach 3.1 Turbo.
And being so adamant about it too.

>>58156230
Oui oui baguette salami cheese si bu flei senoritta

>>58156237

>Juanrga

I bet he posts here too.

>>58156190
According to companies that actually want to make money, yes, that's literally the only purpose for a cpu to exist. The rendering/video making crowd is a meme tier, you either go higher and pander to servers and data centers where you can make millions with just a single deal or you pander to gaymen crowd which consists of hundreds of millions of poorfags who will still add up to a nice sum - You don't go for the 3% of hipster/apple faggots who work with After Effects or Sony Vegas while using MacOS on pre-build hardware that always favored Intel over AMD. AMD CPU department is literally finished.

>>58156284
So much this. Those CPUs are literally worthless and have no consumer base. If they're really this bad when it comes gayming, /v/ermins will stick with Intel. The whole lane up is too weak to compete with Intel server solutions unless there are some 32/64c Zen CPUs that I don't know off. It literally panders to small crowd of 3D/2D rendering faggots that are well know for brand loyalty. Zen is DOA.

>>58156252
:3

>>58154885
>intel 4c/8t beats intel 8c/16t
INTEL BTFO

>>58156403
Not in the same market, they shouldn't be compared..

>>58154811
> (You)

>>58155304
Kill yourself.

File: 1476814367092.jpg (55KB, 600x374px) Image search: [Google]

55KB, 600x374px

>>58156403
>muh games which don't utilize half (maybe 3/4ths if you count the 16 threads) of the processor
>LOOK AT IT GETTING BTFO IN MY SPECIFIC USAGE SCENARIO. AMDOWN SYNDROME

Here's your (You).

File: 1476808569370.png (87KB, 285x330px) Image search: [Google]

87KB, 285x330px

>>58156341

>>58153734
>not running your 955be @4.2ghz for 7 years.

>>58156571
Shhh, don't respond to it, it won't go away otherwise.

>>58154008
Doesn't count

>>58156057
I don't know why so many clinically retarded people like you post here.

The 6700k is faster than the 6900k which costs more than $1000.

The first benchmark doesn't scale linearly with more cores but we can clearly see that the Zen CPU is good enough for gaming even at stock clocks.

The second benchmark shows the 6900k to be 15% faster than the 8 core Zen but it is clocked 8% higher.

We can't know the pricing until it's released. Even if you believe the leak the $500 Zen refers to the special edition and the regular one will cost around $300.

Why do you guys do it for free?, these thread feel like /v/

File: 1476802533329.jpg (12KB, 225x257px) Image search: [Google]

12KB, 225x257px

>>58156651
Because, we are ryzen

>>58156230
What's with all these shitty scans?

File: 1454283744336.jpg (7KB, 204x212px) Image search: [Google]

7KB, 204x212px

>>58156674
ok

>>58155068
Don't know about how their I series scales, but I know an 8 core xeon would see something like 1/1/2/2/3/3/4/5 turbo stepping with a 500mhz max turbo. 8 to 1 core left to right.

Intelfags last week
>M-muh unbeatable IPC wall

Intelfags now after seeing near equal IPC
>M-muh clockspeed

You do know that whatever frequency advantage Intel has will be diminished by a more mature process, or just simply a SKU with less cores, right? What's the next excuse gonna be? Intel's secret weapon they've been stashing away beneath the crust to unleash when things go south?
Just admit that Intel got caught with its pants down like they did with Netburst, they practically did nothing performance-wise for 4 years and are now gonna answer for telling the consumer market to roll away and die every year.

Good riddance.

Fucking AMD fuck them fucking pajeet pooshits cunts with crappy drivers

>>58156206
I'm in the -0.39%,
switched over to an Intel build after a decade of only using AMD

man, I do not know how I could support a company like AMD that makes terribly inferior lower speed & crap quality products

Intel for life for now on, it just works - when it is Intel inside

>>58157171
I lold, finely played good sire.

Does this mean I didnt fuck up by just buying the fx8350?

irrelevant when we dont know price.

>>58157246
You did.

>>58157257
The 8 core should be put to running real workloads, the 4 cores and 6 cores to a degree are fine just running games, these are still really branchy masses of boilerplate and that represents a real-life workload.

>>58157257
what is this screenshot implying?

>>58157287
Well shit. I can always sell it for at least 75% of what I bought it for to some kid on craigslist as the 8 CORE GAMING MONSTER.
I sell stuff on craigslist all the time for as much as I paid sometimes even more.

Why doesnt it turbo up to 4ghz like intel? would fix gaming.

Why isnt there a 120w version clocked at ~3.7-4ghz?

>>58153715
This is an engineering sample and final clockspeeds are not set in stone yet.

>>58153805
And the 2xx series
and the 7xxx series

I'm still on the 6970 because they haven't done a good bang for the buck card yet. 480 is almost good enough, but it had the power connector clusterfuck AND it's only a midrange card.

Anyone else hype for Bulldozer?

I hope there's 6 core chips. More cores dammit!

>>58157371
Because it's a fucking 8 core and even Intel's 8 core doesn't go over 3.7 with a much higher TDP.

Want the 4GHz version? Wait for the 6 or 4 core.

>>58153893
>you get a 8 core for the same price as the 6700k.
thats pure speculation

>>58157310
It was a correct prediction about /̶g̶/̶ /v/.

Perf looks fantastic but AMD will fuck it up in some way or another.

I'm betting the review samples will have turbo disabled because it's bugged and will lose 15% performance, but AMD will enable it in a month or so after everyone and their mother already posted underwhelming reviews.

Fucking AMD

>>58153760
>>58153801
In some circles they are referred to as newfags.

>>58153760
they come from /v/.

>>58157391
>Because it's a fucking 8 core
Not a reason, you can always do it with just 1 core. which is what 6900k does to 4ghz, look it up.

>even Intel's 8 core doesn't go over 3.7 with a much higher TDP
As you can see in the third graph the 6900k only uses 3 more watts. but like i said, no reason AMD couldnt do up 120w.

>Wait for the 6 or 4 core.
Why would I want that, then id just get a intel now.

>>58157421
pretty fucking retarded prediction desu

>expecting anything decent from AMD

>>58157507
It's an ES, we don't actually know turbo clock on the 3.4 base model, it could be 4GHz for a single core as well, or it could be a 3.7GHz for all cores, which would demolish the broadwell.
And AMD doesn't want a 120W+ TDP because these are AM4 boards, they share the socket with the upcoming Raven Ridge and they don't want to make them expensive.

File: 1409692028915.jpg (24KB, 250x251px) Image search: [Google]

24KB, 250x251px

5 years later and there's still no reason to upgrade my 2500k

>>58157654
Other than being a piece of shit with no USB 3 support, no NVMe support, much less USB 3.1 or TB support?

>itt: intel shills don't know what an engineering sample is

>>58157682
They do but they're playing retard at the moment since reality is hard to swallow.
Let it sink in, it is Xmas.

>>58157571
that maks sense, thanks.
I guess they'll make a server line with huge tdp but that will probably be very expensive?

>>58157665
>no USB 3 support
bullshit
>no nvme
fair enough, but not really worth it anyways.
>USB 3.1 or TB
for what purpose?

>>58157748
Server parts will have a lot more cores and lower clocks, bad for gaming.


From what I remember Sandy doesn't have native USB3.0 and uses Asmedia or something controller.

NVMe is quite worth it for the latency alone.

Every new phone comes with 3.1 and TB is fantastic for DAS.

>>58157805
>From what I remember Sandy doesn't have native USB3.0 and uses Asmedia or something controller.
maybe so, but it works well.

>NVMe is quite worth it for the latency alone.
I cant notice 2ms. Im sure its good for some niche usecase tho.

A phone wont saturate a 3.0 connection.
Nor will a harddrive, both TB motherboards and TB drives come at a premium too. Theres plenty of SATA.

>>58155314
The glue doesn't come anywhere near the parts you need to cool though, like you said it's jus there to keep the IHS in place.

It's the fact that they used TIM instead of solder that fucks up all the K chips since Ivy Bridge, and also some i3s as well.

>>58157748
>I guess they'll make a server line with huge tdp but that will probably be very expensive?
Expect them to put out 16 or 32-core MCM Opetrons by middle next year. They desperately want the server market. I'd guess they'll keep the clocks at around 1.6-2.2Ghz, with very high end models going to 2.5Ghz.

>>58157654
>5 years later and there's still no reason to upgrade my 2500k

The dual core 7350k outperforms the 2500k for $180.

>>58157682
Fucking stop you've beensamefaggibg this whole thread about engineering samples and you're still wrong

Kill yourself

>>58157911
that's $70 less over 5 years. Doesnt even cover the the cost of a mobo.

>>58157920
The boost on that chip is the base of what Lisa Su said.

It's a fucking ES Sample from around the middle of this year for whatever company Canard got it from to test AM4 boards. It's not optimized and it has low clocks. It still hits around Skylake in per-core performance in games, and near Broadwell-E in overall performance.

This thing's going to force Intel to change things up drastically, deal with it.

>>58157945
Except on a new mobo you also get USB3 ports, pcie v3 ports, more pcie channels, DDR4 support, NVMe for 2gbyte/sec SSDs, etc...

>>58158022
see >>58157748
>>58157892

>>58157963
>This thing's going to force Intel to change things up drastically, deal with it.

Yeah, they released an overclockable i3.

That's about as drastic as they have to get. They did a low-end but high-performance chip because Zen APUs would be competition in that area. In everything else, they can just say "5% cheaper for every OEM who doesn't order from AMD", and they'll be set.

Oh what is that? EU crying antitrust violations and a 1 billion € fine? That's okay, it's chump change for Intel, and they get more profits by paying that and keeping it up.

>>58157911
>dual core
lul okay champ

>>58158059
You sound like you enjoy Intel being scummy and getting away with it.

Great thread, /v/

>>58157892
>I cant notice 2ms. Im sure its good for some niche usecase tho.

nvme isn't about the latency but the throughput. 2gbyte/sec ssd > 550mbyte/sec ssd.

>>58158066
has hyperthreading, unlike the 2500k, so core count is the same as the 2500k.

>>58158067
I'm just realistic. You can bet your ass that this will happen.

>>58158089
No, NVMe is actually all about IOPS and latency, throughput is nice but there's no reason even USB 3.1 can't have high throughput, SATA is simply shit.

You'll notice a far larger difference going from 100k IOPS to 300k IOPS than from 500MB/s to 2000MB/s sequential.

>>58158089
you know hyperthreading is not the same as actual cores, right?

but keep sucking on intels teet, let me know once its worth it to upgrade from a 5ghz 2500k

>>58153096
The zen chip isn't even running past 3.4ghz

Not to mention no one but AMD has a chip

>>58158059
>Yeah, they released an overclockable i3.
They're also pushing 6-cores into their main line with Coffee Lake (a little too late desu)

Also they did that before, and doing that multiple times would net higher fines.
They might be able to eat a billion dollars (though not easily) but they definitely can't eat 10 or 20 billion.

>>58158138
That's silly, if AMD wants to get anywhere with OEMs they must send them ES'es at least half a year before launch for testing, debugging and validation, the OEMs must prepare motherboards and buy off chipsets.

Saying AMD only has the chip is silly.

>>58158138
see
>>58157963
>It's a fucking ES Sample from around the middle of this year for whatever company Canard got it from to test AM4 boards.

Did the fucking cancer go to sleep or did they automatically abandon thread once it hit autosage?

It's much nicer here now.

>>58153167
Have you done any overclocking with your 2500k?