Zen is DOA

whew, lads
>inb4 engineering sample

>>58278419
That 6900K is most likely running at its actual turbo you faglord. So somewhere between 3.7 and 4Ghz.

>>58278419
translation?

>>58278474

numbers are universal, my anon

>>58278472

so your point is Ryzen needs to turbo boost ~50%? Good save.

>>58278472

6900k's all core turbo is 3.5ghz, the zen es in this test was also 3.5ghz all core turbo (you can see it in the OPN).

>>58278419
Yes, I believe your poorly mspainted chinese crap.

>>58278419
How much and how fast RAM did the AMD machine have? Do the runes say?

>>58278521
3.5 is the 1 core turbo. 3.3 is the all core turbo
same clocks that French magazine had and that tested within 13% of the 6900k. Would be consistent with the 14.69% in OP's runes, whatever they mean.

Yes! Thanks anon, that's a load off my shoulders, I knew that fucking french shit wasn't real, it couldn't be real, this is and proves Zen is garbage like I knew from the start.

>>58278419
I'll wait for the launch. Then I'll be disappointed when it turns out IPC is half of Intel's current gen.

File: amd3.jpg (66KB, 568x612px) Image search: [Google]

66KB, 568x612px

>>58278419
dumb intelfag

File: 1410046519529.jpg (6KB, 250x241px) Image search: [Google]

6KB, 250x241px

>ES
>moonrunes

This is what I just scored with my 6900k clocked at 4.5GHz (AVX -2 multiplier):

MM Int: 967.9MPix/s
MM long-Int: 286MPix/s
MM Quad-Int: 4.32MPix/s

MM Single-Float: 889.47MPix/s
MM Double-Float: 509.1MPix/s
MM Quad-Float: 18MPix/s

>>58278907

4.5GHz/(967.9/724.89) = 3.37GHz

File: 1481407088039.jpg (45KB, 576x694px) Image search: [Google]

45KB, 576x694px

>amd Africa

>>58278419
Why does it say 3.2GHz/4.2GHz for the 6900K? Is it OC'd?

>>58279064

calcs here
>>58278907
>>58278998

put it at ~3.37GHz

Apparent engineering sample running at non-retail clocks, manages to be competitive anyway.

shit bait

>>58278419
SOURCE PLS

>>58279324

>competitive
lolwut? Did you look at the numbers?

>>58278419
>engineering sample
kys

File: 1465864687459.png (28KB, 499x322px) Image search: [Google]

28KB, 499x322px

why would anyone root for zen to bomb?

its like you enjoy being assraped on pricing from no competition

>>58278662
Shitty bait, only good enough to make me reply. Enjoy your (You)

>artificial mememarks

Isn't this supposed to be a $500 chip while the 6900k is over 1k?

File: 1b70c8cd-73d1-4cab-bc44-903e46d0351a.png (105KB, 255x345px) Image search: [Google]

105KB, 255x345px

>doubting this based man

Zen wil be shit like everything AYYMD in the last 10 years.

>>58279879
nobody knows

nothing so far can be considered solid other than what AMD themselves have shown us

>>58280101
>revisionism
f a n b o y
a
n
b
o
y

>>58280118
>nothing so far can be considered solid other than what AMD themselves have shown us
And that should be taken with a grain of salt.

Let's see some prices before calling it DOA

File: spzZZ3q[1].gif (4MB, 320x180px) Image search: [Google]

4MB, 320x180px

>people actually believed amd's lies and fudged numbers
>again

does anyone know how much the ISA extensions (say, AVX) affect these tests, and how different are these CPUs in that respect?

>>58279053
Arica

>ryzen
What kind of 12yo came with that name? It's like they asked someone in a twitch stream chat. What the fuck? How can you take amd seriously after this.

>>58280951
>fudged numbers
We saw real-time benchmarks friendo.

>>58281015
>Arica
Africa without a single f given

>>58281143

yeah you saw a video stream of benchmarks without any investigation into the underlying hardware, code for said benchmarks, or any other pertinent data. It's called marketing, boi, and was done for a pump and dump of their stock.

ZEN IS A FLOP
AMDKEKS ON SUICIDE WATCH

>>58281221
So you're telling me when they were having e-sports players playing and streaming at the same time, it was staged all along?
woaaaah...

>>58281259
Oh shit are you saying that its fast enough to play a game?

Sure hope so.

Fast enough to compete with Intel? Nope.

>>58281259

I didn't watch that one - did they show someone remove the cooling solution to show the actual chip packaging? Did they compare that actual game with the i7-6900k benchmark-wise? If you can play games on it for """e-Sports""" you sure as hell can show some real-time third-party-verified benches with it.

>>58281259

any cpu since core2 can encode stream-quality h264 video (720p, 60 fps) in real time

>>58278474
No, trust the numbers goyim

>>58281221
The blender benches straight up couldn't have been faked, they released the versioning and the file they were benching with.

S O U R C E

>>58282588
No, don't you understand? It's impossible for AMD to create a chip that is comparable to Intel's chip because it's AMD! I'm a smart person.

File: c285f51235dfefd70f6ecec16283c8b0.jpg (129KB, 669x1193px) Image search: [Google]

129KB, 669x1193px

>>58278419
DOA?

>>58281305
prices are everything in competition friendo

File: Dankest_78ef86_5907611.jpg (44KB, 398x960px) Image search: [Google]

44KB, 398x960px

>>58283033
checked

>>58283752
only poor people care about price

>>58283803
End your lonely existence.

File: 5175051+_404d9413f1533fda5fc07f6e344e060d.jpg (28KB, 493x386px) Image search: [Google]

28KB, 493x386px

>>58283803
>not being frugal
>not saving money if at all possible
>"IT'S 2% BETTER! WHO CARES IF IT'S 69 TIMES THE PRICE OF THE NEXT BEST THING?"
We don't even have any solid information yet, and won't until they start selling it. Trying to decide anything before is clearly just baiting or jumping the gun.

Enjoy living paycheck to paycheck, wagecuck.

File: gradoducks.jpg (38KB, 500x401px) Image search: [Google]

38KB, 500x401px

Never trust Chinese leaks lads.

>>58283876
>Never trust Chinese
ftfy

>>58283803
Ateasl I got money in my pockets, unlike (You).

What does pic try to say?

>>58278419
>twice as expensive cpu doesn't perform twice as good
Intel BTFO

>>58281015
Anywhere from "not at all" to "orders of magnitude" depending on the workload you provide. This test appears to be one that mostly tests AVX (my jap isn't good enough to see what other things are tested), which mostly affects applications coded specifically for AVX and those compiled with optimization flags to take advantage of AVX (but even so, the app needs to have an easily parallellizable workload for AVX to have an effect).

To my knowledge, AVX adoption in consumer applications has been pretty poor. Games won't be affected by this, and neither will most other applications. The exception is highly specialized applications, or apps with easily parallellizable workloads, but you'd need to code them accordingly and compile with a compiler with support for the AVX version in use to take advantage.

I assume that "DOA" means " Death Of Amd" (?)

>>58284208
Dead on Arrival

>>58281355
Assuming they wern't using the hardware based solution in the GeForce expereance, that takes a lot more load off the CPU so you could easaly do 1080p @ 60fps on a core2.

>>58284229
Thinking Zen will ever arive, AMD will be bust befor then.

>>58284055
That would explain it. Ryzen is not good on AVX. There was some instruction throughput chart that shoed this.
Guess the reasoing is "GPUs are better at this anyway, why don't you do it on the GPU?"

>>58278482
not all numbers are equal.
you want to tell me that intel is twice as fast as amd after amd has shown real world performance fine, ill believe it, but I need to know if any of this is real world performance of it it's all synthetic best case scenarios.

granted, these best case scenarios are why I honestly believe amd won't charge 500$ or more for the base model.

At least this is how I see it.

>>58284292
GPUs do AVX and SIMD really really well if it's not very branchy code, but some stuff is still just going to be relegated to the CPU because of how it branches.

>>58284319
In this case, AMD is still only about 70-80% of Intel's performance for FPU operations.
They've pushed that up from around 55%, and their integer performance is nearly on par, about 85-95% now, and in rare edge cases they can beat Intel.

>>58280199
>>58280118
the french leak can be considered solid as they are VERY transparent with it and actively answered questions relating to it.

Its still an engineering sample but is a very good bare minimum expected performance.

>>58281015
>>58284055
>>58284292
Integer performance is also low in OP. There is just something wrong with the test setup in OP.
Does Sandra record actual CPU speed?

File: sandra-multi.jpg (80KB, 601x793px) Image search: [Google]

80KB, 601x793px

Performance figures for the i7 6900k in the OP are entirely fake.

An i7 6950X clocked at 4ghz doesn't even score over 600mpixels.
Looks like a desperate little intel fanboy through this together trying to disparage AMD knowing most people are too fucking dumb to look up bench results themselves.

intel shills are terrified of Zen.

>>58284387
>most people are too fucking dumb to look up bench results themselves.

http://ranker.sisoftware.net/show_run.php?q=c2ffcee889e8d5e3d7e6d3e2d0f684b989afcaaf92a284f7caf2&l=en
http://ranker.sisoftware.net/show_run.php?q=c2ffcee889e8d5e3d5e3d1e2d0f684b989afcaaf92a284f7caf2&l=en

OPs results are real.

Question is: what did they do to the engineering sample? Not consistent with any other leak.

>>58284476
>OPs results are real.
No, no they are not.

An i7 6900k at 3.2ghz does not perform better than an i7 6950X at 4ghz.
Not all submitted results are valid, you shit eating idiot.

>>58284496
6900k results range from 394.98Mpix/s - 779.50Mpix/s. There are a couple very close together at 394-405 , but the 569 posted in OP is already the 8th lowest score.

How does Sandra get such a wide range of results?

>>58284668
Much like Cinebench its an incredibly easy bench to spoof, and it does a poor job of reporting actual clockspeeds.

>>58279747
Because it will make a lot of people on this board really really assblasted.

>>58279747
Because they're paid by intel.

>>58284387
>N-no this can't be happening! DELET

Fucking kill yourself you AMD nigger.

>>58284696
/g/ is designated shitting board for ayymd and MS poojeets for quite some time

>>58284711
Nice shitpost, kid. Stay in /v/ where you belong.

>>58279747

Shorting AMD stock

>>58284790
working overtime poojeet?

File: rb8PUC-h.jpg large.jpg (30KB, 480x234px) Image search: [Google]

30KB, 480x234px

>OPs results are real.
Wut?
>legit benchmarks
>ssl error
If your benchmark suite isn't willing to spend $50 on an SSL cert why should I trust it?

>>58284858
chinks still run xp and IE6, newfag

File: thumbsup.jpg (8KB, 236x214px) Image search: [Google]

8KB, 236x214px

>>58284872
>re-enforces my point

>>58278419
ryzen 3.14ghz

nuff said

>>58284887
>take my balls SO deep

SINGLE FLOAT VS INTEL

HAHAHAHAHAHAHAHAHAHAHA

HOLY SHIT THAT CANCER PERFORMANCE

DEAD ON FUCKING ARRIVAL

LITERALLY

BULLDOZER FX REHASH GARBAGE

File: 0e7.gif (220KB, 280x199px) Image search: [Google]

220KB, 280x199px

>>58278419
lel they copied the score from the first column into the third. Can people really be this gullible?

>>58285734
row*

>>58285734
No, that's a result of the processor itself not being on the market; There's only 2 results available for that particular Zen ES, both of those benchmarks gave identical scores, hence the score range being what it is.

http://ranker.sisoftware.net/show_run.php?q=c2ffcee889e8d5e3d5e3d1e2d0f684b989afcaaf92a284f7cafa

>>58284346
newer SIMT handles coherent branching OK, and future generations seem to be focusing more on wavefront/warp shuffling to keep ALUs occupied.

CPUs will dominate simulation workloads near-term, but that's not necessarily guaranteed forever.

Lmao

If AMD somehow managed to make a CPU that outperforms a 1000$ Intel CPU, why would they sell it for 500$ or less? What would stop them asking for at least 800$?

>>58287063
When AMD had a leading product, as in dominated (almost) all categories, they charged more than Intel top-for-top product.

And they should, and will, price accordingly.

If AMD doesn't push 8 cores in the mainstream market Zen will be DOA.

>>58287063
Because the $1000 CPU will almost certainly drop in price once there's some real competition.

>>58287063
Lmao

LMAO

>>58287091
So there are 2 things that can happen

a) It's shit like originally expected, but reasonably cheap. But that doesn't matter if it's not good enough

b) It's on par with i7 extreme chips, but costs just as much, making it unaffordable (except for enthusiasts who most likely have a 6-10 core intel already)

File: totally legit guise.png (290KB, 1500x714px) Image search: [Google]

290KB, 1500x714px

>Sandra
>a benchmark that only ever gets recompield when Intel puts out a new architectur
>somehow remotely related to AMD

That and (as has been said) its easy to spoof results - much like cinebench.

>>58284273
Saved this.

>>58284346
>In this case, AMD is still only about 70-80% of Intel's performance for FPU operations.
We knew Zen was going to be short in FPU based tasks for a few weeks now, the CanardPC info just helped confirm it. Intel's FPU is wide as shit. Even then, these are all things we're seeing from ES chips and validation mobos, which for all we know is throwing errors out the ass. Need a full report on the testing and results ideally other than just benchmark numbers.

>>58287131
Zen will honestly probably do a lot better in the 4c Raven Ridge APU platform.
Workstations have been a declining form factor for going on a decade.

> but stealing Intel's lunch of fat Xeons margins is the biggest immediate goal

>>58287260

>he didn't get the memo

All engineering sample leaks always represent the final product if it makes AMD look bad. If Intel ES leaks look like shit then well its not final silicon so it will be fixed y'see?

>>58287287
I don't give a rat's ass about a 4 core Zen. I want atleast a 6 core affordable Zen CPU.

>>58287330
>>he didn't get the memo
Must have been that new Mexican kid down in the mailroom. Fucker is always losing shit.

>>58287334
you, me, and probably a lot of /g/ are at least interested in Summit Ridge, but I'm pretty sure we're in the global minority.

>>58287091
literally the last time they shit all over intel with their product, they had one sku that matched intel's 1000$ one, priced it at 300$ then everything better was 1200$ and 1500$ with a 20/40% faster cpu respectively.

amd, though their history have made skus that attempt to make you look at their compeditor and ask "Why bother" rather than "Well, they are both the same price" because if they are the exact same price, as in 100% feature parity, preformance parity, and whatever + or - they have in either of those are proportionally lower in price too, everyone would go intel over amd.

consumers because all the fuckin intel commercials, and pro segment because they are not a shit company to work with and devil you know over devil you don't.

not only that, but this cpu is literally a sweet spot cpu.

The most performance, for the lowest cost, smallest size, and they are only making one non apu sku of it. Everything more is mcm, everything less is cut fucked up dies, only making one sku makes the production cheaper as you don't have a 2 core sku that needs to do well to make up for lesser 4+ core sales. not to mention amds 8 core at 350$ would pull 300-400% more profit per mm then the 8350, and die size dicates cost more then what intel did.

>>58287334
Well, its looking like AMD is going to try and fill that market. Given that Intel's 6 cores costs twice as much.

I think AMD is going to completely sidesweep the market with another bang for buck processor.

>>58287131
amd is going to push 8 core as the mainstream standard with 4 core 8 threads as the minimum.

>>58287375
yes and no, the majority IS interested in it, but the majority also doesn't know what it wants.

8 cores is so much better then 4, but there is no benchmark that will show you why, you have to use it to get it. there should be a load balancing software, something like prime 95 but it hits the cpu in different ways to simulate shit just started and is fucking with what you want to do, kind of a synthetic way to show off load balancing and what it brings to the table.

>>58287440
the benchmark is:
> run a couple browser tabs in background
> let JS adware botnet hell go on as per usual
> try to get something else done if foreground

the real growth in CPU demand the last 10 years has come from how bloated webpages are getting.

>>58287391
I guarantee you, if they release a bang-for-your-buck 6 core Ryzen that actually can compete with Intel for once all the waitfags/poorfags will go and get one. Then once they gain alot of market share it'll only get better from there.

>>58287199
or

c) amd cut enough features out or down so that 80% of the time, its parity with the 8 core and only made one sku so cost they need to recoup from development is cheaper, along with getting people on an amd based motherboard so a future cpu upgrade or two is capable.

Im honestly willing to bet there are quite a few people who would get a skylake or something similar if intel didn't swap out socket compatibility so fucking often.

>>58287440
A 350 dollar Zen isn't mainstream.

>>58287440

Unless price is astronomical i'm going to be all over the 8c/16t chip despitei t being enormous overkill for my current needs. I have no plans to get into cpu heavy tasks that can into threads but that much horsepower means i'll basically never have to upgrade this side of 2025.

Plus I can always boinc on the side.

>>58287494

According to a few chumps at OCN (and here to be honest) zen needs to be significantly faster than Intel and half the cost at twice the thread count because otherwise why not just get Intel?

>>58287503
that's the high end of the mainstream prices whether you like it or not.

>>58287474
web pages, web page bloat, that makes up a bit,

browsers being shit at their jobs is another thing.

>>58287516
same here, I largely have no real need for the full power of an 8 core 24/7 but when I do something that wants that power I damn sure would fuckin like it, if intel didn't sell theirs for 1000$ i would have likely gotten an 8 core sooner.

>>58287539

Fat zen, 32gb fast DDR4, beefy cpu cooler (I really do hope I can get a bracket so my AM3+ cooler fits, if not its not a huge loss), quality motherboard and later on full fat Vega will have me set for 4k (and basically anything else I can think of) for quite a while.

Hmm, seti@home loves gpus and I do want an orion slave girl...

>>58287564
>(I really do hope I can get a bracket so my AM3+ cooler fits, if not its not a huge loss)

Depending on the socket arrangement, might not even need a bracket. I'm more than confident that my 212EVO will just go right onto AM4.

>>58287091

This.

People here think AMD is their savior. AMD is going to line their pockets just like Intel if they have the chance.

>>58287516
>significantly faster
That won't happen. That's why AMD needs to undercut alot, sell their chips more than just half the price of that of an Intel equivalent.

>>58287608
AMD isn't the savior, competition is, but only AMD is able to provide any against the incumbent Intel at present.

Also, AMD at least hasn't had the opportunity to pull as many jew maneuvers as Intel in the past, so they're not as instinctively hated.

>>58287494
honestly, there are a few factors that I need to know and its an instant buy

1) what is gaming performance like compared to intel's 8 core

Like it or not, I don't have money to separate my work machine from my hobby/leasure machine
If it keeps up, perfect

2) I need to see where amd fails. make no mistake, amd is not going to be better than the 6900 in every task, I need to see where it fails, and how often I would push it to fail with my normal use.

3) Assuming 1 and 2 are met, I need to see what this auto oc shit is, because this is the difference of me getting a base 8 core or the higher end 8 core if the difference really is binning and it will automatically set itself to the max the cpu can do given my cooling solution.

amd potentially has me for 350$, possibly for over 500 based entirely on the usability of the cpus and what the boost is.

>>58287091
Yeah, it's kinda funny and pretty sad how people think AMD are the "ebin good guys" just because.

>>58287603

Pic related is my cooler. I'm not worried if I do need to replace it (would be a good excuse to go for something like the D15 or R1 ultimate) as i'll keep the fans so meh, its not a total loss should it play out like this.

>>58287609

Ah but for previously mentioned chumps it has to be cheaper AND faster. If its cheaper and performs the same? Then just go Intel!

(there really is no logic so these people, they want AMD to compete just so they can buy Intel for cheaper which clearly shows no interest in AMD at all)

http://www.overclock.net/t/1619110/cpc-first-unofficial-ryzen-benchmarks/660#post_25746977

This is the sort of idiocy that fills /g/ and what AMD has to fight.

File: silver arrow.jpg (19KB, 300x300px) Image search: [Google]

19KB, 300x300px

>>58287715

Woops forgot pic.

>>58287663
I can tell you with confidence that the "auto overclock" as we've been badly calling it is a more advanced implementation of their AVFS set as used in Excavator and some hardware implementation akin to Intel's current style of turbo boost, nothing more. It isnt so much an "auto overclock" as it is a finer grained control of core boost states than their current Excavator chips offers.

If not manually OCing you will still be limited by the binning process VID table hardcoded into the CPU along with the hardcoded safe temperature limit controls, which themselves are limited by the hottest current piece of the CPU under load.

A much much better cooler might let you see 4Ghz+ on low core/thread count or light load applications, nothing fantastically crazy.

>>58287728
Looking at the amount of play in the mounting plate it should be ok. Not 100% sure though.

>>58278419
b-b-b-but this is the weakest model!
the final one will be 100% faster and it will be free!!!444

>>58279747
people hate amd shitposters, fanboys, and paid posters. they are the fucking worst. like a child throwing a tantrum, every single day in every fucking thread.

so yeah, that is one good reason.

>>58288041
>people hate shitposters, fanboys, and paid posters. they are the fucking worst. like a child throwing a tantrum, every single day in every fucking thread.
ft.
fy.

>>58287715
>(there really is no logic so these people, they want AMD to compete just so they can buy Intel for cheaper which clearly shows no interest in AMD at all)

The person you linked to isn't wrong though. Amd isn't able to demand a premium price, because look at intel, pretty much every single 6900 can clock to 4.3 stably, and because so much shit is unchanged, you know what you are getting into, if amd was lets say 200$ cheaper, Id go intel just because I know exactly what i'm getting.

In this case, amd either needs to have a significant edge over intel to go in early on, or has to have the better value to the point that even if there is a bug found that is bad, it still was a good deal due to cost.

>>58287760
If I remember right they mention the cpu has no upper limit, it all comes down to what your cpu cooler can handle.

I want to see more on this, but honestly beleive it's just turbo with many states, but imagine if it wasn't.

>>58288028
actually, this one is sub their weakest model, at least shit post accurately.

>>58288041
wasn't nvidia the only one that has paid posters in threads? were other companies proven, or are we just assuming?

>>58288270
Unfortunately I have no faith in AMD's financial and marketing department to price Ryzen 'right'.

>>58288421
Su gutted AMD's marketing department when she got into CEO. And no doubt will know exactly where to price Zen to get it sold. She's doing work. Don't have faith in AMD, but have faith in the Keller/Su Tag Team

>>58288421

>Unfortunately I have no faith in AMD's financial

Its not like they tripled their stock value in 12 months or anything.

>>58288827
>$1.90 to ~$11.50/$12
>trippled
m8

>>58288859

AMD confirmed for bankruptcy - anon said it so it must be true.

AMD's stock has been and will probably remain a roller-coaster ride.

>>58287440
>amd is going to push 8 core as the mainstream standard with 4 core 8 threads as the minimum.

summit ridge isn't mainstream and AMD isn't going to sell an 8 core chip for $350.

>>58290070
yes, their entire 4 6 and 8 core lineup is not mainstream.

>>58280064

Absolute pussy destroyer.

>still shitposting about an engineering sample
>>>/v/

>>58291155
I think you're just misusing "Mainstream" in place of "Consumer" here, or "Most popular" Statistically speaking, based off of current processors sold, their Quad Core would be their "Mainstream" product, while being in their "Consumer" line. While the Octo Core could well be the most "Popular"

>>58278419

>Running the ES at 3.14Ghz locked
>Running the Intel chip at 4.2Ghz turbo

Filthy japs shilling for Shittel

>>58284809

This guy gets it

Buy long term puts soon

>ITT: BUTTBLASTED AYYMDPOOJEETS

>>58284858
>paying for SSL when letsencrypt offers certificates for free
Seriously there is no excuse.

>>58291155

amd's current fx lineup is HEDT, just priced dirt cheap because bulldozer wasn't competive. they're also not 4/6/8 cores, they're 2/3/4 cores. their mainstream offerings are the APUs which top out at 2 cores.

>>58291390
but you want them to do well so intel products will be cheaper

>>58283803
>only poor people spend money needlessly

>>58280064
wasnt the leader of team a female desu

>>58279747
AMD shills are cancer and overhype their company's bullshit to death

They deserve everything they get

>>58278419
>AMD fails to deliver yet again

WHAT A SURPRISE, COULDN'T HAVE SEEN THIS COMING FROM MILES AWAY

>>58292392
you dont understand the difference between zen and the construction cores do you?

>>58280064
I heard he repopulated an entire midwestern town into thriving metropolis with just one cumshot.

>>58279747
>butthurt about cpu prices
there are other more important price tags in life to be concern about besides being able to run mobas at 144fps, /v/irgin

>>58292392
AMD chips unironically have 4 6 and 8 cores. They share only the FPU between 2 threads. In Intel CPUs 100% of the CPU execution units and resources are shared between 2 threads.

>>58296038
I would never have purchased an $1100 6900k, but 8 cores at Sandy Vagina or better IPC at >=3.4 GHz for <=$500 is a must buy for me this Spring.

gg, asian mom & shit-wrecker.

>>58296340
What makes you think you'll be getting that for 500 bucks? AMD isn't a charity.

>>58296038
>amd chips unironically
Sure. They're underpowered in some ways, but they're cores.
>Intel shares between 2 threads
Sure, but Intel calls that hyperthreading, they don't call it another core.

>>58296369
That's the point, though. It's not another core if the resources are all shared like in Intel CPUs. AMD only shares the FPU, so these are legitimately individual cores.

If AMD had 100% resource sharing like Intel, their FX chips would be better performing than Intel chips in single core integer workloads because they have 1 extra execution unit.
If AMD really had 4 cores on the 8 core FX chip, it'd be a strong CPU.

>>58296355
currynigger rumors. :)

8c pure CPUs are frankly of comparable size or smaller than newer 4c + garbage graphics APUs that sell for $300 anyway.

I expect most Summit Ridges will just be binned samples that don't hit the golden perf/Watt marks for the target frequencies for Naples (<=2 GHz for 4-chip MCMs), where AMD likely expects to sell by far more Zeppelin chips.

>>58296038

theres only one decode unit per '''module''', which leaves one of the two 'cores' bottlenecked as fuck and unable to execute instructions.

>>58296932
I didn't even know they had only 1 decoder. That's insane. So in effect each FX core has 1/2 of the decode capability of an Intel core.

>>58296988

pretty much. they didn't fix it until XV, so each of the cores in a piledriver CPU end up underutilized unless you feed it a very simplistic workload like 1 million MOV instructions.

>>58279053
>African Made Devices

>>58294667
Yep. Suzanne Plummer.

>>58297774
The second decode was added in Steamroller.

>>58296355
what >>58296436 said along with even at 350 it would be close to 400% proffit uptick over the 8350, and they showed it off at a public event to gamers.

amd is dumb as fuck some times, but would they have made a public gaming event that was completely revolving around a cpu more expensive then most gamers entire systems?

I mean, is it a secret that intel has been ramming us in the ass for the last... what, 6 years on price? they dropped die size 3 or 4 times and still charge the exact same for damn near the exact same performance.

The way I see it, there are enough flaws in amds offerings, along with enough concessions for making it cheaper that it could never demand a 500$+ price at base unless there was something truly special there. And no, just matching intel in some areas is not truely special.

>>58288345
Nvidia was just being just and righteous, sending missionaries into this desert of a board, preaching to the red scum, bringing the light of true brand loyalty and gaming performance to them.

>>58279747
intel shills

To be honest, if i can pay $180-200 for something within 5% of i5 performance in gaming i'd be happy.

>>58300198
This.

People seem to believe that in order to compete they must BEAT Intel performance, when in fact all they need to do is get close to Intel performance for a much reduced cost.

It's all about the performance per dollar.

>>58300384

Yes, and I'm very perssimistic about AMD pulling this off. Let's hope they can actually do this.

>>58279747

being shills

we need more cpu brands than Incel and Ayymd, and more GPU brands than Ayymd and Novidya

>>58300860
>Who are Matrox
>Who are Via

As much as you may wish it to be so, there will never be more than there already are. Via, Intel, and AMD are the companies that hold the keys to the x86 kingdom, and they're more than happy to not sell cuts of those keys to anyone else.
Same with the GPU market, Matrox, nVidia, and AMD have it locked down with patents, there are no ways in.

>>58300911
>Matrox
>Via

literally who?

3 competing firms is enough, but Matrox and Via should maybe expand the market to Europe, never heard of them

>>58278419
Ni Hao, Xian.

20 Intel Quick Computing Points™ have been deposited into your account.

>>58300935
Because Matrox and VIA cannot compete. VIA are still on 40nm, with clocks up to 1.6Ghz.

And Matrox haven't been able to produce their own GPUs for years, they're forced to license AMD GPUs, slap them on a custom PCB, and then sell them to enterprise only.

They're not consumer products.

>>58301011

so, basically trash

well i guess i have to hope that incel, ayymd and novidya don't become turbokikes

>>58278419
Confirmed DOA but it will sell like crazy in poor countries (asia/africa).
Stay intel boiz !

>>58300384
I say this alot, I buy amd not because I like amd, but because I despise intel and nvidia for business practices along with fucking me over, if amd is a viable option, which for me they always have been both in graphics and at the time of purchase cpu, ill go amd.

cpu wise, If i had to buy one today, it would be intel because amd just is not an option, zen, even if it fails will be an option as its at minimum a 40% uptick average over excavator and that puts it in the relm of good enough, with some loads/uses surpassing intel's high end.

>>58300384
Yes I agree, but AMD should still have "beats all" product for brand image. It goes a long way for the consumer to say "If you want the best you get Nvidia or Intel" even if it's in no way cost effective.