It looks like AMD fanboys overhyped another piece of trash yet

I'll believe it when I see the actual benchmarks performed by an mostly unpartial mostly trusted party.

>>57736807
If your gunna shill Intel at least post the source

>>57736807
>it's ok if intel does it

>>57736960
Y u still exist?
Answer me.

>>57736807
That site is absolute shit. No sources whatsoever.

>>57736988

Intel does not false advertise though.

If Intel says something about one of their products it's probably true.

>>57737036
the source was a wccftech article that pulled a claim out of its ass that it was on par with 6850k

actual AMD demo puts it with 6900k
it is shit

>>57736988
Yeah because they're true, every ad about them beating AMD is true ;)

>yibada
>That Intel processor wasn't identified in the press release, but has been assumed to be the Core i7-6900K Broadwell E.
>The report claims that the 8 core variant of Summit Ridge will have performance equivalent not to the 6900K, but to the 6 core Core i7-6850K.
>That certainly has a different flavor doesn't it? In effect, each Summit Ridge core is only about 75% as fast as a Broadwell E core.

>AMD Summit Ridge Processors With $250-$300 US Pricing To Have Performance Equivalent To Intel’s Core i7-6850K

>$250-300 summit ridge
>75% performance for 50-60% of $500 6850k (sale price)
>75% performance for 42-50% of $600 6850k (regular price)

>overhyped piece of trash

2 shekels have been deposited into your account >>57736807

>>57736807
If this were true, the rumored $350-$500 launch prices for the 8c models wouldn't hold up.
If Zen falls outside of 85%-95% of Broadwell IPC I'd be surprised.

> will obviously not get quite so close in perf/W though, especially compared to SL/KL

>>57737165

and just for good measure, when it comes out and everyone buys one up resulting in no stock leading to stores jacking up the prices much like they did with the rx 480 we'll get a bunch of threads with,

>$500 summit ridge
>75% performance
>amd BTFO

>>57737113
>actual AMD demo puts it with 6900k

actual amd demo didn't specify what cpu it was paired with, those shitty "journalist" sites you read did that

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

71KB, 552x661px

>>57736807
DELET THIS

>>57737165
AMD nor fanboys have ever claimed that Zen will beat Intel's HEDT
Anyways, Intel's actions speak tons, they're pushing 6 core chips into the mainstream, releasing unlocked i3's, and making 28 core Xeon's
Competition is already on in the entire market Zen covers
>>57737228
>If Zen falls outside of 85%-95% of Broadwell IPC I'd be surprised.
They're at parity IPC wise (as long as you don't go into 256 bit FPU ops), clock wise everything points to a disadvantage

When is Zen coming out?

File: IMG_20160713_195647.jpg (300KB, 768x1024px) Image search: [Google]

300KB, 768x1024px

>>57737337
right now

>>57737337
rumors say January, but who the fuck knows?

>>57737301
>>57737236
>>57737228
nobody really knows where the IPC will fall outside of cherrypicked benchmarks, but most pundits expect it to come closer to Haswell/Broadwell than Sandy Bridge/Ivy Bridge.

>>57737262
Intel is making moves on the possibility that Zen will be good, not with any solid knowledge.
They could easily stick jew prices on all those new products if Zen underperforms.

>>57737337
Around January/February, they will probably make big announcements and demos at CES

>>57736807
The intel CPU is clocked 33% higher.

>>57737337
OEMs have zen now

i work at dell we have a few systems to be released for Christmas but its more of a stealth launch.

>no actual source
Just desperate short sellers who lost all their money betting on AMD to lose.
Even if Intel is 30% faster per core, AMD is offering 50-100% more cores for the same money.

>>57737519
Cool. My dad works at nintendo. Can we be buddies?

>>57737519
Laptops or desktops?

>>57737525
The i7 Broadwell-E chips are currently on sale ($350/$500 for 6 cores at least), so the perf/$ comparisons are somewhat in flux.

I just want Summit Ridge to actually expose the on-die 10 GbE controllers, which would add a shitton of value and start eating Xeon-D's lunch.

i have never been more happy to have jumped the AMD ship

>anally raped at every price point in the GPU market by Nvidia
>facefucked by Intel at every price point in the CPU market

How is this company still even around? What the hell happened to AMD? Five or six years ago they used to be actual competition to Intel at least with GPUs. Now its just flop after flop.

AYYMD IS FINISHED & BANKRUPT

File: amada.jpg (80KB, 853x1024px) Image search: [Google]

80KB, 853x1024px

can we get some new amada content for zen release? yes?

>>57737615
But the first point is false and the second point isn't true in the OEM market and in the budget market
Why are you blatantly shitposting?

>>57736807
>what now, /g/?
This
>>57736923

I hope Zen is bretty gud. I'll buy one even though my 4690k is still doing awesome

File: 1461604786839.png (18KB, 744x247px) Image search: [Google]

18KB, 744x247px

>>57737695

>>57737545
Cool contact info?

>>57737547
desktops

no laptops due for another 4-5months at least.

>>57736807
So this 8 core chip is going to be slightly worse than a 5820k? Was hoping to upgrade to 8 cores but I wouldn't have much reason anymore. Maybe the server chips are worth the wait.
Been using my 5820k for over 2 years now, I'm glad AMD buyers will finally be able to buy purchase the level of performance that has been available for years at the same price as It has been for years.

>>57737615
>How is this company still even around?
early 2010's, they mostly got carried by the bitcoin market and the good will of their fans. also by selling their stuff at meager profit margins to remain somewhat competitive. a lot of their efforts in the past 5 years or so were spent on R&D, to prepare for polaris/vega and zen.

lately, huge ass SoC contracts with console manufacturers and chinese hardware vendors have helped them a lot. in addition, polaris is doing extremely well despite nvidia fanboys trying to convince themselves otherwise.

>What the hell happened to AMD?
a whole bunch of shit

in no particular order: getting way too optimistic and building a bunch of fabs (that are now globalfoundries), spending a fortune to buy out ATI, gambling on parallel computing and losing nearly all their x86 market share, selling APUs that nobody wants, and putting out consistently terrible marketing efforts. the company was steered wrong, pure and simple.

of course, there's the corporate sabotaging by intel, but that's relatively minor in comparison.

>>57738223
>at the same price as It has been for years.
you mean half the price

>>57738256
8 core is 350, friendo

>>57736807
Sad!

>>57738280
Many such cases!

as long as the price is half of what intel does they'll be good

>>57738267
8 core, 16 thread. go look at how much the 5960x costs.

if the IPC truly is only 75% of broadwell-e then that's 75% of the perf for a third of the price

>>57738240

I disagree on the marketing bit. Even with the shit hand that they've been given (i.e. the products they have to market ...) they've done a real great job by guerrilla marketing (by which I mean basically lambasting Nvidia and to a lesser extent Intel on twitter and in interviews etc) and stirring up the reddit and the kids on 4chan type of crowds. Everyone seems to love an underdog, and especially when they stoke the 3.5GB memes to boot.

>>57738301
The only thing I can see as a benefit is extra pcie lanes. Still unconfirmed if the $350 chip will have 40 lanez or whatever, or if that is only for the $500 part. Totally underwhelming.

>>57738319
That is only as strong as a 6 core. I.e. a 5820k

>>57738240
>they mostly got carried by the bitcoin market
Bitcoin mining was profitable on GPU's only during the 6xxx era, after that ASIC's hit the market and killed it, Litecoin/scrypt mining only lasted the first few months after Hawaii was launched
>also by selling their stuff at meager profit margins to remain somewhat competitive
The 290X was 700 USD at launch

They did mostly scrap Bulldozer on the HEDT and enthusiast market to develop Zen, but the GPU division hasn't stopped launching products (Fury series is right there) and they just launched new excavator APU's in September
And they are competitive with Intel's offerings in the OEM market
>in no particular order: getting way too optimistic and building a bunch of fabs (that are now globalfoundries)
The fabs were doing well before AMD's CEO back then decided to spin them off and name himself CEO of them
>spending a fortune to buy out ATI
Literally the only market where they are competitive and the only reason they have big OEM contracts right now
>gambling on parallel computing and losing
Zambesi/Bulldozer wasn't intended just as a parallel architecture, at first it was intended as a serial and parallel monster while reducing area as 2006 AMD slides showed, there are even ES samples floating that had a much different silicon than Bulldozer
They scrapped it around 2009 since it didn't deliver in reducing area at all, and then everything became a clusterfuck
>selling APUs that nobody wants
APU's have been successful all around, it's the only product where they compete with Intel and they secured them multi year long semi custom massive contracts

File: reluctant chuckle.gif (3MB, 286x258px) Image search: [Google]

3MB, 286x258px

>>57737247

File: kokoro3.jpg (353KB, 800x600px) Image search: [Google]

353KB, 800x600px

when will we see Zen powered Anime Processing Units?

>>57738656
>inb4 Amada is scrapped for Zen APU's launch

>>57738443

Not that guy but:

>Bitcoin mining was profitable on GPU's only during the 6xxx era

That's still a fairly long time, and a hell of a lot of chips would've been sold.

>The 290X was 700 USD at launch

Conversely, the amount of time that it was anywhere near 700USD was very short.

>but the GPU division hasn't stopped launching products (Fury series is right there)

Which hasn't been competing well, at least in terms of revenue or even average cost per board.

>and they just launched new excavator APU's in September

Which nobody really actually *wants*, and the remainder that do - Aren't paying enough for them to generate a good profit. The same goes for the OEMs that buy them because they're cheaper than intel's CPUs (which really, are APUs also). That cheapness, or more accurately the customer perception of cheapness when it comes to AMD's APUs, translates to a lower price per chip that the OEM is willing to pay, and thus less profit per chip. The volumes might look big, but compared to the ridiculous write-offs on the massive over-stocks of older APUs they've already lost heaps on, and the R&D costs of making the damn things for the little margin they'd make ... Eh, it's not anywhere near the cash cow it was supposed to be.

>>57738833
>That's still a fairly long time, and a hell of a lot of chips would've been sold.
Bitcoin wasn't popular during the 6xxx series, most people didn't trust it at all
I was around back then, a 6990 would pay itself in 3 days, I didn't buy it because I thought it would die out and be replaced by some other coin
>Conversely, the amount of time that it was anywhere near 700USD was very short.
Only due to the 780Ti launch which was priced quite lower than the 290X, meaning the 290X was competitive with Nvidia's
>Which hasn't been competing well, at least in terms of revenue or even average cost per board.
But it has been competing well in revenue and cost m8
>the R&D costs of making the damn things for the little margin they'd make
They're made on a 5 year old 28nm node, with rehashed Bulldozer which has nearly no improvements and with graphics IP derived from the Radeon division
They're exceptionally cheap to make, just like the Polaris 10 dies

Can someone point me to benchmarks comparing the i7-6850 and 6900?

>>57738937
Why not the 6800?

>>57738937
>i7-6850 and 6900
http://www.anandtech.com/show/10337/the-intel-broadwell-e-review-core-i7-6950x-6900k-6850k-and-6800k-tested-up-to-10-cores/8

>>57739013
Of course all you fags care about is gaming

File: 6e1447ff284ab7626e07184942612e76.jpg (76KB, 665x498px) Image search: [Google]

76KB, 665x498px

>>57738827
Amada was already scrapped awhile ago

that or she has been removed off the internet and only exists in rare Jap merchandise form

Why can't and ever seem to compete with Intel, is Intel just that good?

File: wow.jpg (14KB, 480x293px) Image search: [Google]

14KB, 480x293px

>>57739013
>entire performance spread less than 1%

>>57739055
>Amada was already scrapped awhile ago
I'M LITERALLY SHAKING
This was the only good thing AMD marketing has done in the last decade wtf I hate MAD now

>>57738240
im really shocked how prevalent are the marketing memes.

>>57738910
It was popular enough that AMD were having supply issues and the prices of the cards were skyrocketing as a result.

Maybe only a few people were into BTC, but they must've been buying 4+ per person. In fact I heard there were some groups that bought dozens for severfarms.

Actually now that I think about it, the shortage continued well into the 7000 series man, what are you talking about? By the time the craze fully ended, the 200 series was almost out.

>>57738910
The point stands that they were only selling it at a *good* margin for a very short period of time, and even then when the 780 Ti came out, the price instantly went to similar, and then quickly fell from there. In other words, the price slid rapidly because it wasn't very competitive at equal pricing, and certainly was nowhere near competitive at its original price.

>>57738910
Lol no?

http://www.anandtech.com/show/10613/discrete-desktop-gpu-market-trends-q2-2016-amd-grabs-market-share-but-nvidia-remains-on-top

https://jonpeddie.com/press-releases/details/add-in-board-market-increased-in-q316-amd-lost-market-share-while-nvidia-ga

AMD have been bleeding market share for ages now - Except in the last year they've been gaining back a bit of what they've lost (this isn't something praiseworthy, it just means that the losses temporarily dipped into 'disastrous' before settling at 'really bad').

And even then in the last quarter they've once again slipped in share - They fell to ~20% market share last year, and even before this quarter where they've started to slip again, they never regained their losses.

But you only need to take a look at the actual revenue charts - Even with less market share, Nvidia was still gaining on AMD. How? The ASP was ridiculously higher than AMD's - In other words, by going with the premium 1070/1080, they were giving loads of low end customers to AMD, only to make a hell of a lot more money by dominating the high end - The advantage is clearly Nvidia's.

File: article_147485976957e892f9af91f0.png (1MB, 1200x899px) Image search: [Google]

1MB, 1200x899px

>>57739414
Now it's time for Mi hoshi ai.

>>57738910

>They're made on a 5 year old 28nm node, with rehashed Bulldozer which has nearly no improvements and with graphics IP derived from the Radeon division
They're exceptionally cheap to make, just like the Polaris 10 dies

I'm talking about the entire cost from day 1.

There would've been significant TDP challenges to deal with, not to mention the cost of developing updates like Steamroller etc etc, all arches which never ended up in a pure CPU form, which means the R&D cost layed squarely on the APUs. And like I said, even with how cheap they were, they massively overestimated demand and lost many millions on unsold stock. That only happens due to incompetence or a bad product.

>>57736807
Broadwell E = $450 to $1800
Zen = $150 to $500

75% performance for 33% of the price, that sounds like a win to me.

>>57739485
>It was popular enough that AMD were having supply issues and the prices of the cards were skyrocketing as a result.
That was during the 290X days, which died out in months since scrypt ASIC's were launched pretty fast
>Maybe only a few people were into BTC, but they must've been buying 4+ per person. In fact I heard there were some groups that bought dozens for severfarms.
You don't know jack shit at all, the people that actually splurged in 290X bought dozens at a minimum
>Actually now that I think about it, the shortage continued well into the 7000 series man, what are you talking about? By the time the craze fully ended, the 200 series was almost out.
There weren't massive shortages during the 6xxx and 7xxx days, the real shortages were during the 290X launch
By the time the 7950 was launched ASIC's had dominated the market
>The point stands that they were only selling it at a *good* margin for a very short period of time, and even then when the 780 Ti came out, the price instantly went to similar, and then quickly fell from there. In other words, the price slid rapidly because it wasn't very competitive at equal pricing, and certainly was nowhere near competitive at its original price.
The original price was due to literally no competition at it's performance bracket until the 780Ti was launched
>AMD have been bleeding market share for ages now - Except in the last year they've been gaining back a bit of what they've lost (this isn't something praiseworthy, it just means that the losses temporarily dipped into 'disastrous' before settling at 'really bad').
During the Kepler days they were getting marketshare back
>>57739536
>I'm talking about the entire cost from day 1.
Most R&D had already been covered by older releases, they also bear a fraction of R&D cost compared to Intel's since they don't have their own fabs
>There would've been significant TDP challenges to deal with
Most power-related IP is applied as well in the Radeon division

will there be massive clearance of stock of old fx cpus in january then?

>>57738910
>Bitcoin wasn't popular during the 6xxx series, most people didn't trust it at all

I bought a 6950 and launch, the one with the "cut corners" on the PCIE connector, that could be flashed to 6970.

4 months later I was unfortunately out of job, and I made a modest living out of mining for half a year. If I didn't need to cash in every 3-4 weeks, I'd be swimming in dough today (made 4-5 BTC a day back then).

When the ASICs came out, I preordered one for something like $200, it cost $300 when they finally released it, and it was already completely fucking useless by the time they shipped it. Made something like 0.01 BTC a night. Luckily, I managed to sell it for $350 two weeks later, so I made more money off of it that way than actual mining. I feel kind of sad for the poor sap who bought it... and more sad for the people who bought them for $400 and $450 a month later (when the difficulty already quadrupled and they made even less money).

>>57739654
i would imagine, you could prob make a fx 6300 builds for days. cant imagine what you would use them for.

>>57739714
>I bought a 6950 and launch

*at* launch. 2010 december I think? It wasn't any better than the 5850 I had before, but cost twice as much for some reason.

>>57739588
Idiot

>>57739395
http://www.anandtech.com/show/10337/the-intel-broadwell-e-review-core-i7-6950x-6900k-6850k-and-6800k-tested-up-to-10-cores/6

>>57739719
im almost 4th world so for my new pc

>>57739844
>4th world

Where do you live then?

>>57739897
On planet deviluke.

File: Jack_Kirby_'s_Fourth_World_Vol_1_1.jpg (92KB, 400x617px) Image search: [Google]

92KB, 400x617px

>>57739897
>4th world

He probably lives on Apokolips.

>>57739897
poorest part of poorest yuropoor cunt

>>57740011
oh, greece?

>>57740080
something like that just without sea

>>57737236
>actual amd demo didn't specify what cpu it was paired with, those shitty "journalist" sites you read did that
are you fucking retarded

https://youtu.be/oQS8s7TOXsE?t=1m25s
>On the audience's right, is an Intel Broadwell-E Core i7-6900K CPU

File: rA0clo3.png (805KB, 1440x2560px) Image search: [Google]

805KB, 1440x2560px

Kek

>>57740551

AMD finally embraced the meme.

TFW: FX 8320 @ 4.93Ghz with H80i GT and RX 470 with core undervolt -40Mv

Setup is dead quiet because the weather has been comfy lately (22-26°c i live in place with an overall 35°c+ temps from 9am to 11pm)

I have chill in my arms, my legs are a bit too warm and my nutsack is at the comfiest of temperatures. Thanks Amada!

Im thinking on removing the H80i and install my old TX3 during the summer to heat my entire 2 floor 4 rooms house.

>>57740833
It doesn't matter what cooler you use, retard.

>>57740884

>H80i is designed to cool 140W+ CPUs.
>Keeps my FX 8320 under 59° with stress test.
>TX3 desinged for 95W chips.
>FX 8320 is 125W + a 1.4Ghz OC and .3V Core bump.

>>It doesn't matter what cooler you use, retard.

Are you even trying?

>>57740833
Post cpumark scores

File: 14479564_10205805371411749_4147202251520129354_n.jpg (66KB, 960x472px) Image search: [Google]

66KB, 960x472px

>>57740928

W8 for me to get home. @Work right now dude.

I dont benchmark tho so you'll have to wait a couple hours. (5Pm i clock at 6PM and drive 30mins to home)

I will use a fucking tag o u know its me.

pic related is an old pic i've posted a few times here.

>>57740921
H80 is loud and is no better than air cooling

>>57740996
Sweet. I don't think I have ever really seen a heavily OC'd 8350 on passmark.

>>57741000

Well since the last update i created custom fan curves and outside BF4, Vegas 14 and CAD the thing is quiet as fuck.

The RX 470 its a bit loud when im doing Vegas 14 renders and im using a Nitro+ (sometimes i regret not going for the MSI one).

>>57741057
I was kinda leaning towards the XFX one. Good to know.

>>57740921
I was talking about heating up a room

File: amd_zen_logo.png (16KB, 306x246px) Image search: [Google]

16KB, 306x246px

>"Zen will shit on the competition"
What did Pajeet mean by this?

>>57736807
At least it'll drive prices down, there's no point buying "enthusiast grade" hardware.

Can we ban retards who post wccftech """""rumors"""""?

>i7 6900k is over $1200 CAD
being a leaf is awful sometimes

>>57736807

This like the first C2D's, 2nd gen will dial it in.

AMD is the moral CPU

File: 1474186595806.jpg (43KB, 440x374px) Image search: [Google]

43KB, 440x374px

FEMALE CEOS ARE A HELL OF A DRUG

WHO /IN AGREEMENT/ HERE

we've known for (literally) months now that zen only performs on par with sandy/ivy bridge, amd has a history of making EXTREMELY exaggerated claims about performance in marketing material (bulldozer, fury x overclockers dream, 480 beating 1070, etc) and fanboys had only themselves to blame for falling for it AGAIN.

zen might be a good sidegrade for people who don't want to be stuck in intel's jew trap of buying a fast i3 or a slow as fuck i5, but otherwise it won't be as good.

>>57742355
>AMD is the moral CPU
Would have agreed if they didin't include fucking hardware botnet in their cpus. They can go bankrupt for all I care.

>>57742530
nani?!

Intel literally has hardcoded backdoors and telemetry

AMD is completely free of that shit

>>57742552
>AMD is completely free of that shit
>this is what blind amdpoojeets actually believe
https://libreboot.org/faq/#amdpsp
AYYY LMAO

>>57742576
o fugg thts spooky

>>57742506
The engineers working on bulldozer were extremely conservative with the numbers they put out. Marketing and the really dumb fanboys pushed the dumb ideal of it being a godly chip.

Fury x was literally never said by AMD to be the overclockers dream.

480 was also never shown to beat 1070. 2x 480s were shown to beat a 1080 in aots, though.

>>57743191

hello pajeet, denialism won't change the truth

>>57743216
I'll give you a real (you) when I get home.

Maybe in the meantime you can actually source your claims.

>>57737093
>*increase* the power of your CPU by typing a code
>totally not just unlocking a feature you were cucked away from in the first place

File: 17-1080.573265681.png (251KB, 1920x1079px) Image search: [Google]

251KB, 1920x1079px

At equal clocks Summit Ridge matches comparable 8 core/16t Broadwell-E in Blender.
SiSoft Sandra results look pretty good too.

What AMD claimed is a 40% uplift in IPC over Excavator, and they later stated that Zen cores have the same energy per clock as Excavator.
A look over at the Anandtech bench tool will show a binned Carrizo(Excavator) Athlon vs an i5 2500k
http://www.anandtech.com/bench/product/1684?vs=288

The Athlon here doesn't have L3, so performance would be marginally higher, but this is the best baseline to look at. In some metrics Excavator is only 15-20% behind Sandy Bridge with pretty close clocks. Other metrics, which tend to be FPU heavy, its further behind as expected because of the narrow FlexFPU.

An average 40% uplift would put Zen about on par with Sandy Bridge in some things, and around Ivy Bridge to Haswell in other metrics. Obviously the example we see in the Blender demo is a best case scenario for the arch. Also likely is that AMD's arch is displaying slightly higher multicore scaling, or slightly higher SMT throughput vs intel's arch. IPC in this workload still could be lower, but the combined throughput of all cores when utilized makes up for it.

On the clock scaling side of things, we know that the Excavator module in Carrizo has this power curve:

XVR with AVFS
3700mhz: 22.5w~
3500mhz: 16w~
3150mhz: 10w~
2800mhz: 7w~
2450mhz: >5w~
2100mhz: 3w~
1750mhz: >2.5w~

10w per core for 3.15ghz.
Not too long ago the ID string for an OEM Summit Ridge sample leaked online. Some OEMs contacted Dresdenboy and confirmed it as authentic. The string shows a 3.2ghz base clock with a 3.5ghz turbo.
8 cores at 10w per core +15w for uncore works out to 95w very nicely.

Stop falling for fanboy hype.
This is the most realistic and factual post on Zen you're going to see on this board until NDA is up and reviews come out.

>>57739588
Words cannot describe how fucking stupid you are

Buy (((Intel))) goy, you really need that extra CPU power for looking at trap porn.

>>57740226
>believing AyyMD
Those results are exactly in line with an underclocked i7-6800K, NOT the 6850K or the 6900K. In fact, if you look up the average time it takes for the 6800K, you'll note that it's a few seconds faster than the time noted in the demo.

I bet you also think that Cinebench offers a legitimate comparison of core architectures

>>57743697
>At equal clocks Summit Ridge matches comparable 8 core/16t Broadwell-E in Blender.
No it doesn't. That demo was faked, they either turned off two cores on the 6900K or they flat out lied and used a 6800K. Anyone with a 6800K can vouch that the stock six-core Broadwell-E is significantly faster than either of the times shown here.

>>57743920
Stop pulling things out of your ass.
The demo was heavily scrutinized months ago when it was relevant, and no one found anything. They even released the render scene they used, and people scoured through the code of their Blender builds. The only tiny bit of controversy that emerged was one tech illiterate retard on Twitter finding an L1 cache loop and thinking that it was a sign of AMD cheating. Blender themselves came out and stated that this is just how the program operates.
It was 8c/16t Summit ridge at 3ghz vs 8c/16t Broadwell-E at 3ghz.

You're a delusional shitposting fanboy desperate to fit your underage /v/ tier console war narrative.

>>57743985
>The demo was heavily scrutinized months ago when it was relevant
No it fucking wasn't
too many journalists were too busy sucking off Lisa Su's dick to question their results. Meanwhile, the Broadwell-E community were scratching their heads wondering why the "downclocked" 6900K took so fucking long. The only way those times made sense if if the 6900K in the demo had two cores turned off to make it into a 6800K with lower clocks, and even then the 6800K doesn't take that fucking long to finish the benchmark.

File: 1478739371426.png (202KB, 416x500px) Image search: [Google]

202KB, 416x500px

why get Zen when you could just buy a broadwell E right now?

It's not like Zen is cheaper or anything, $500 for an 8 core thats barely on par with an Intel 6 core

75% of broadwell-e
multiplied by 8

all depends on pricing for me

>>57744102
source me on that price nigga.

>>57737247
fanboyism for computer parts. Now I've seen it all.

>>57744222
u must be gnu here

>>57744248
nah this all started when console kiddies decided to buy computers. Its asinine and retarded. trying to start a platform war within a platform. I know this is going to sound hypocritical but this is nvidia's fault for trying to push shit like gameworks and destroying performance in games they get their hands on. I cant get 60 frames with the new skrim se with a rx480 because of this bullshit.

>>57744308
Brand fanboyism has been a thing even when 3dfx was here so not sure what you're on about

>>57744354
I wasnt here that long ago. there must have been a very long lull because this only just recently started happening since I have been here.

>>57736807
Fuck off intel shill

>>57744049
>pulling even more things out of your ass
>outright denying reality
>doesn't know about the anandtech article on it and subsequent forum discussion

Yep. Butthurt childish fanboy.
My case stands.

>>57744407
Here specifically? Then you're still wrong. I've been browsing this shithole since 07

File: 1478537204045-pol.png (102KB, 360x432px) Image search: [Google]

102KB, 360x432px

>>57736807
>not waiting for actual benchmarks
>being this retarded not recognizing this as Intel trading blows via a paid off shill

>>57741375
>What did Pajeet mean by this?
He meant that Zen will be able to render 3D Poo models faster than intel can, but it will suck at everything that isn't rendering poo models.

>>57744442
What a dumBfcuk. I hAve no words for how delusIonal you are. There arE chilren in North Korea who are less Delusional than you.

>>57744513
>p-please stop calling me out for my sourceless fanboy claims
Immature console war fanboy mentality belongs on /v/. Stay there among your own kind, kid.

>>57744464
nope.

File: 81875.png (38KB, 650x450px) Image search: [Google]

38KB, 650x450px

People should be buying the 6850K because they need the extra cores, because when it comes to real world usage , the extra cores end up not making any difference.

No one is going to buy Zen if its only strong point is multi threaded shit and failing at the things people will do it with everyday , like game with it.

AYYMD HOUSEFIRES

>>57744566
That bench is showing a heavy GPU bottlneck.

>>57744566
>muh gaymen!
Intel shills are getting desperate

File: untitled-12.png (23KB, 687x694px) Image search: [Google]

23KB, 687x694px

>>57744602
better ?
the result ends up being familiar, the extra cores don't do much for gaming

Where they shine are in cinebench or similiar applications that make use of all the cores.

>>57744707
>muh rendering!

>>57737337

January for their high end

March for budget

>no Poof
>base of it is a rumour
Lmao at your life.

>tfw the ZEN hype is already priced into the share price.
Time to sell?

>>57744407
>inelite
>1.7
>woodscrews

Is 2010 really that recent?

>>57744566
>GPU bound frame results all within margin of error

(Lenny face)

>>57736807
140 replies

ctrl+f "wccftech" "archive"

no URL/source

dropped

>>57744729
Are you a retarded mongo?
>muh gaming
What part of GPU bottleneck don't you understand?
You can't compare different cpus in an environment where you can't use any of the processors to their full potentials.
Let me give you a game analogy so your 79 IQ brain understands...
benching cpus based on games is like benching gpus based on game install time.

try benching firefox compile times.

>>57744845
>wccftech
POO

>>57744909
>GPU bottleneck
That clearly isn't the case with all the CPUs tested there.

>>57736807

Love those fucking sources. Wow.

Seeking Alpha once said my company would go out of business within a year. That was four years ago.

Fucking idiots.

>>57745153
Seeking Alpha is 100% stock manipulation central.
They exist to manipulate small tech stocks and try to profit from rallies and short sales.

>>57745153
>my company
Post yfw AMD Pajeet is confirmed /g/ lurker

File: bf1-cpu-benchmark-dx11.png (41KB, 705x441px) Image search: [Google]

41KB, 705x441px

>2017
>dual core
dual cores processors need to die.

File: 3 KB.jpg (5KB, 203x212px) Image search: [Google]

5KB, 203x212px

Does no one like the idea of putting one of these in an HTPC or something? I mean... they are generally pretty cheap, right? Sounds like something I could be tiny-hype for.

>>57745247
thats the thing they will be cheap and probably do just fine with pretty much anything entertainment you could want. I mean lets face it all AAA games will be made with very little cpu use in mind.

>>57743682
But it still increases performance doesn't it?
Thats not false advertisement

>>57744909
>try benching firefox compile times.
is this what you people in in your spare time ?

>>57737337
Q1 is amds answer
rumors put 8 core at january with the 6 and 4 at march.

>>57745247
>Does no one like the idea of putting one of these in an HTPC or something
Ahem
>no iGPU
>high thermals for a small, lo-airflow enclosure
>power-hungry compared to Intel's Apollo Lake and Skylake-T
How about no?

>>57745811
It wasn't like I was ready to fight IRL over it. Just asking.

>>57738443
once the lite coin happened, you couldn't get an amd gpu for a good 6 months without severe price gouging.

as for mining,

when polaris came out, it allowed mining on the gpu to be profitable for a small time.

Its all about gpu power and gpu power use, and amd has a range of 90-160 watt for polaris 480's

On the amd ceo, before he went to GF he signed a retarded contract with them... he fucked amd harder than anyone/anything else ever has.

as for buying ati, it gave them gpu tech, which they integrated on the cpu... most of us have discrete gpus, and turn our noses up at apus, but fuck me, being able to dump a cpu into a motherboard and it works does wonders for low end users, and seeing its usually twice the speed of intel until intel decided to dump dram on the cpu for it, Id say its a success.

if they put 1-2gb of hbm on the cpu, you will see a huge performance across the board increase for amd, low end laptops will never be the same again.

>>57738833
little brother bought one day 1 and got it for 650.

as for gpu not competing... has there been a whole lot of need? I mean look at it this way, look at a 290x and look where it lands today

also, remember fury got fucked over so god damn hard due to the 20nm process falling though and caused amd to just buy out the contract. the furry competes, but due to the larger process, it doesn't compete well (not to mention nvidia stockpiling failed titan x'es only to sell them the moment amd made the announcement for the same price. grate marketing on nvidias part but dick move all around)

and those write offs are all thanks to the prior shit bag ceo and retarded contract with gf

>>57739485
the old miner craze did, its why you could get a new 7970ghz for sub 300$, but then lite coin happened and whoops, every mid range + card is now double the price for 6 months.

>>57745811
>>57745875
It might still be cool for a 1080p60 FPS Gaming HTPC

I doubt it would stay relevent for more than 12 months though

>>57739013
>>57739795

>>57739048

Because In our leisure time, most of us play games. Not sure about everyone else, but god knows I like to not think and just get engrossed in something that is completely skill based after a long shitty day.

The bonus for more then 4 cores comes in doing other stuff on the side, along with the overhead.

not to mention if you use your computer for work, you got a best of both worlds scenario, good gameing, good productivity.

>>57738910
you know those claims intel makes? that 20% faster and such, bulldozer was making actual improvements like that, granted the cpu was such shit it meant dick all grand scheme for single core, but what's currently made is nothing like original bulldozer performance wise.

>>57744102
there is a good chance, unless they stated otherwise, that zen+ will fit into a zen socket, so upgrade path for one.

For me, I have considered intel for a long time and some deals, if they were within 400 miles I would take them, but anymore, the wait is also bringing the prices for parts i'm going to get with it down too so i can justify the 1tb ssd, and possibly a 500gb ssd boot for the other oses in the same cost as a 500gb ssd use to be when i considered intel.

my computer is good enough to not require an upgrade, so i'm lucky there, but fuck me would an upgrade feel damn good right now.

>>57744566
now have windows update start in the background, and take out an entire core, or an automated process startup that is cpu heavy.

I know have your web browser shit itself and damned several cores worth of processing power till it un fucks itself.

This is where more then 4 cores comes into play at least on the gaming side of things, and with modern apis, all cores are more efficiently used so more cores does offer better performance.

File: 1474301761268.png (6KB, 415x416px) Image search: [Google]

6KB, 415x416px

>>57737834
I love this.

>>57736923
>an
i hate you

>>57740833
the fucking fx6300 and my 3rd world 970 kept me cozy all winter. Why would you ever not want a housefire computer?

>>57739596

>That was during the 290X days, which died out in months since scrypt ASIC's were launched pretty fast
>There weren't massive shortages during the 6xxx and 7xxx days, the real shortages were during the 290X launch
By the time the 7950 was launched ASIC's had dominated the market

Wth are you talking about? Shortages were rife during the middle/later half of the 7000 series.

Look - Bottom line is that there was a very substantial period of time (6000 series to 200 series) where the sales of AMD cards were heavily boosted by coin mining (bitcoin or litecoin or otherwise). That type of random boon is unlikely to happen again, and inflates sales figures during that time. In other words, AMD got lucky.

>the people that actually splurged in 290X bought dozens at a minimum

You're making my damn point.

You said BTC wasn't popular - Granted, only a small proportion of people were into it, but as you've just reinforced, the people that were into it did buy many cards. Again, this was a once in a lifetime random lucky windfall boon for AMD.

>>57739596

Why cherry pick the original price? The 290x was out for such a ridiculously tiny period of time before the 780 Ti came out that the sales figures would've been terrible - Who was going to buy a 290x before waiting for the 780 Ti to come out? Who would be stupid enough to not wait for comparative reviews? ... Only the people who would've bought a 290x regardless of whether it was better or worse. Fanboys in other words.

What's far more significant is the rapid price collapse for the 99% of its lifetime after the 780 Ti came out. That shows it couldn't compete at all.

>During the Kepler days they were getting marketshare back

Uhh, no? It was 37.8% at Q1 2012, at Kepler release, and 35% at Q1 2014, just before the 900 series hit, which demolished AMD's sales overnight.

>>57739596

>Most R&D had already been covered by older releases, they also bear a fraction of R&D cost compared to Intel's since they don't have their own fabs

For the first gen or two of the APUs, maybe. They still had to make everything work in one chip with a TDP far less than their combined components. And from there they were developing newer arches. Also: Maybe the initial outlay would've been less because they were fabless, but Intel makes ridiculously large returns on those fabs, even with the initial costs. That's why their chips sell for such a large premium, and in such large volume. It's also why, as I keep mentioning, AMD keep over-estimating demand and writing down the value of their backlog of APU stocks ...

>Most power-related IP is applied as well in the Radeon division

You misunderstand. I mean the issue is with getting a >60W CPU and a >60W GPU, and somehow making a <90W APU out of it without destroying performance. The end result is the top of the line chips are lacklustre.

>>57736807
He has no source, in addition it wouldn't matter, AMD will still more than likely sell more cores for less than intel

>>57745980

I agree that it's not bad at all to have a cheap choice for basic builds with an integrated GPU - The thing is that Intel also do these APU chips. The entire i3 lineup is basically APUs.

It's only when you look at the chips that are below all of intel's in cost that AMD have something that beats Intel (if only through lack of competition) and those are so low in cost and the market would be so low in volume AMD couldn't be making significant money off them. Just like the consoles.


>as for gpu not competing... has there been a whole lot of need?

... Are you saying picking the higher performing GPU doesn't make sense? That isn't a good thing to do? ... Lol wat.

The Fury didn't compete not just because of the silicon, but because the Arch, like all of the preceding arches, was ridiculously power inefficient. That's why it needed its own bundled cooler, and that's why it needed HBM memory (which is much more power efficient) - And yet still had issues even with that assistance.

As for the write offs - No, the CEO didn't personally predict sales figures, their marketing etc etc teams would've done that. And the contract with GF might've been retarded in other ways, but ultimately only AMD is responsible for picking what amount of APU chips to order, and they did so very poorly.

>>57745186
They have been shitting on AMD for a couple of months now. My guess is that they got caught with some short sales that cost them money around the start of the 480 hype. The butthurt is real. I picked up AMD stock at 2.16 (account didnt get funded when it was down to 1.80) per share. I have tripled my money, but I'll wait until February or March before I sell and move that money elsewhere in my IRA.

File: aa7.png (655KB, 1920x1200px) Image search: [Google]

655KB, 1920x1200px

>>57737649

The shitty source is seekingalpha, which insists that the leaked ashesofsingularity benchmark shows that Zen is somehow bad.
If you actually run the numbers (and believe gaming benchmarks are accurate) it shows that Zen has "muh IPC" somewhere in between Sandy Bridge and Haswell in a benchmark that does not scale beyond 8 threads and therefore does not even utilize SMT.

>>57743920
Did you downclock it to 3GHz and disable the 4ghz turbo?

>>57747938

fuk'n saved arigatou, desu

>>57736807
I just want a moderately acceptable GayyMD rig
I'm tired of being an intel/nvidia turbocuck

>>57742530
yeah this. amd is retarded could have been freedom cpus and shit but no.

>>57745309
>lets face it AAA is straight up shit "movies" for last 2 years

FTFY

>>57737834
That period was the real death of /sp/ btw

>>57748227
I doubt they think the market size of people who would care is big enough to justify making a separate line of non-botnet cpus. Honestly they're probably right.

>>57743872
If it wasn't for ridiculous JS bloated websites and H265/VP9 first gen C2D's would be enough for pretty much every single use case
Prove me wrong

>>57745456
The developers of firefox surely do

>>57744472
>imbliying
It looks like someone who got butthurt for shorting AMD stock
>>57744845
Why would you need a source for such a bullshit article?
>>57745242
Dual core CPU's are perfect for mainstream usage
>>57745247
Wait for the APU's
>>57747007
>You said BTC wasn't popular - Granted, only a small proportion of people were into it, but as you've just reinforced, the people that were into it did buy many cards. Again, this was a once in a lifetime random lucky windfall boon for AMD.
You can't even distinguish between Bitcoin and Litecoin, why are you pretending to know anything?
>>57747050
>They still had to make everything work in one chip with a TDP far less than their combined components.
This is not how TDP works on those chips, they have much less CPU modules than the FX parts, they actually don't have to figure out much regarding energy consumption
>I mean the issue is with getting a >60W CPU and a >60W GPU
There's literally no APU that had the same GPU configuration as a 60W GPU die
>>57747159
>The entire i3 lineup is basically APUs.
The entire Core i* lineup are APU's, bar the i7E's
>>57748275
>implying it's about market size
We all know it's the NSA mandated botnet

>>57747159
on the write off, the old ceo signed a deal where amd would ALWAYS buy at least X number of chips regardless of demand.

again, look where the 290x lands today on charts, the 290x is routinely above a 780ti, and when used in a modern apu, above a 980, and there was one i believe it was also above a 980ti. then the fury and fury x that the rest of what the gap in price was, with the fury x in some benches being above a 1070

you either put out mass market cards like the 460-480 and 1060 or you go for a very small portion of buyers like the 1070 or 1080. with amd going for hbm2 for their gpus, why not wait and make damn sure its good opposed to rushing it like nvidia did? they got mass market, they have driver team working on shit which further increases all cards performance. Its hard to say going after the high end when you can secure a strong base as being worth it.

File: 1435671983407.jpg (68KB, 700x700px) Image search: [Google]

68KB, 700x700px

>>57736807
Noooo, how could this happen, fellow AMD bros?
I invested one year worth of savings on AMD.
I'm sure that article is just fud, AMD would never lie to us, r-right?

>>57738240
>lately, huge ass SoC contracts with console manufacturers and chinese hardware vendors have helped them a lot
their only profitable line of business

>polaris is doing extremely well
division is still hemorrhaging cash

>What the hell happened to AMD?
They bought ATI for too much money.

The best thing that will come from Zen's release is lowering the prices of Intel's competing products. It will not be better than Intel's offerings nor revolutionary in any way. It will be good enough for budget sales.

>>57739414
>we Micro Advanced Devices now

>>57736807
>some chinese click bait site
>makes claims about Zen performance
>based on pootech and jewalpha

>>57749789
>The best thing that will come from Zen's release is lowering the prices of Intel's competing products

You are 100% delusional.

Can anyone link or tell me a guesstimate on the performance of single thread usage for AMD Zen?

I don't care about 8 cores.

>>57750117
Its a 40% IPC uplift over Excavator.
AMD later clarified this was single thread.
Essentially look at Excavator, remove CMT penalty, add an L3, and increase IPC by 40%.
See: http://www.anandtech.com/bench/product/1684?vs=288

>>57749686
but muh youtubers and streamers market!
/jk they send them those cards for free anyway

File: 1471396486391.png (200KB, 578x691px) Image search: [Google]

200KB, 578x691px

>>57747938
Pajeet blesses you

>>57736807
So basically, it'll be equivalent to Sandy Bridge?
lmao

>>57736807
how to get a shitton of (you)'s: the post.

File: Speccy - 25NOV2016.png (18KB, 465x425px) Image search: [Google]

18KB, 465x425px

>>57736807

iT'S A GOOD THING I BROADWELL-E

File: 1419206769559.jpg (328KB, 810x587px) Image search: [Google]

328KB, 810x587px

Huh, didn't know motley fool owned Intel stock as well.

>>57754541

It is considering you got a better cpu than anything from the skylake range.

>>57736807
>It looks like AMD fanboys overhyped another piece of trash yet again

What a fucking surprise.

AMD is shit and their fanboys are pure cancer.

>>57754557
They regularly criticize Intel also. Just read an article where the author suggested firing the whole dev team basically if Intel wasn't going to ship every cpu on a 10nm node by 2018.

>>57754641

So its just Nvidia they shill for?

>>57754598
>I came here to confirm my bias
>a leak with zero credible sources! perfect!

>>57755117

Its the bread and butter of /g/.

>>57737093
2x increase in performance!*

*per watt

I thought zen ipc was on par with haswell

>>57736807
if this is true, what exact intel model would be the equivalent to the 350 zen one? and for what price?
i7-6700? i7-6700k?

>>57755428

The people who actually know aren't saying shit. Far, far too many internet experts have declared Zen a failure because they cannot into understanding what AMD has actually claimed.

The best one is AMD have stated +40% IPC over exavator. So naturally the internet used the latest athlon as a benchline (which AMD didn't) and then declared it fail because who the fuck cares about an L3 cache?

File: AMD-Zen-benchmarks.jpg (55KB, 874x531px) Image search: [Google]

55KB, 874x531px

I'll just leave this here.

>>57750117
amd likely won't beat intel, get your i3 and be happy with it now then.

>>57756020
is it bad?

>>57736807
AMD would get sued to hell and back for telling investors they're comparing one cpu and using another.

>>57756020
but how many cores and how many megamertz?

>>57756222
as fast as a 10 core Sandy Bridge @ 2.8Ghz.
if it's the 8 core Zen it's pretty good depending on clock rates.
If it's a 10 core Zen it's anyones guess.
If it's a 12 core Zen AMD is bankrupt and finished

>>57756305
I'm not sure tbqh. A quite reputable german website claimed it was the 3. 2ghz 8core. The website I got the image from said it can't be known. ¯\_(ツ)_/¯

>>57756451
Oh it could be that they clocked it to 3ghz for comparison but I need to reread that shit I'm drunk.

>>57754557
>we'll catch up in the next gen, promise!

t. AMD, every year ever

>>57756494

Its not like AMD has beaten Intel before. Oh wait.

>>57756451
>>57756490
At 3.5GHz it would be exactly the same IPC as Sandy Bridge
At 3.2Ghz it would be 9% more IPC than Sandy Bridge
At 3Ghz it would be 16.6% more IPC than Sandy Bridge
At 2.8GHz it would be 25% more IPC than Sandy Bridge

>>57756549
you cant afford Intel? what are you POOR? hahahahah pc master race!

>>57736807
So, it performs about as well as a late generation core2 or early i-series?

Typical AMD bullshit. Can't make a decent GPU, can't make a decent CPU.

>>57756647

You deserve a 4chan pass for such quality posting. Assuming you don't have 10bux of course.

>>57756549
>WE WUZ ATHLON 64 N SHEIT

Yeah, "before".

>>57756549
They just compete with Intel hopefully getting the prices right.
>>57756630
Yeah I'm getting that. And I know the 8core zen has a base clock of 3.2ghz and boost of 3.5ghz according to guru3d.
It's confusing to wrap my mind about the german article bc they mention another thing when amd run a blender render comparison between a 6900k vs 8core zen, both clocked at 3ghz. The Zen was claimed to be faster but no numbers or proof. The authors requested details three days ago.
W/e I don't rly care looking that shit up wasn't that entertaining as I thought. I'm out.

>>57756690
silence!

peasant.

>>57756832

>The Zen was claimed to be faster but no numbers or proof.

It was from a test AMD ran on stage - the zen chip beat the 6900k by a second or so rendering an image of the zen logo (that little artistic swirl). There was no hard information given about the systems beyond both chips being 3ghz and the rest of the setup "being identical". Naturally we also have no idea if the build of blender AMD used has been tweaked or not.

>>57756887
Well then it's fucking nothing.

>>57756935

Even if AMD did stack the deck against intel catching up to a 6900k is no easy task.

>>57737519
how is it then???

File: summit-ridge-guru3d.png (125KB, 725x173px) Image search: [Google]

125KB, 725x173px

Then there's this about prices..

>>57757197
is there release date hidden somewhere

>>57756887
Didn't Intel debunk that AMD benchmark as being outdated older version, and if they used newest version who uses AVX Intel was something like 50% faster ?

>>57757284
>It is expected that ZEN Summit Ridge based 8-core processors will launch in week 2 of January.

>>57756020
whats the point of a cpu for 3d rendering when GPUs do that just fine

>post yfw it's all Zen is good for making it Bulldozer 2.0

>>57757394
Yes. Which is why any AMDrone who uses that silly benchmark as "proof" that Zen isn't shit is full of it

>>57757394
yes anon, because Intel is the most honest business out there, please use the updated version "optimized" for Intel :^)

File: 1398748870993.gif (120KB, 252x252px) Image search: [Google]

120KB, 252x252px

>>57756681
>I shape my world views on news headlines.

File: 1479593430253.jpg (229KB, 627x720px) Image search: [Google]

229KB, 627x720px

>>57736807
Looks like the unemployed CTR shills are now shilling for Intel.

>>57736807
75% of 1 single core perf × 8 cores = 6 cores worth of perf

Which is why AMD's 8c / 16t is competing against Intel's 6c / 12t.

Looks like they'll be offering a an i7 3770K on a modern platform for $150, which is cheaper than a used 3770K, currently running about $220

>>57759777
performance does not scale linearly with core count

shitty post undeserving of trips

>>57759830
Depends on the program, moron. No one buys a 4c / 8t part for its single core performance

>>57756630
IPC does not fluctuate with clock speed.
You have absolutely no idea what you're even saying.

>>57757394
No. Blender has had the ability to use AVX for years, and it wasn't omitted from the Blender demo AMD showed.

>>57757982
Stop desperately grasping at straws, you fanboy child.

when are we getting more competition in the CPU/GPU brands?

I love the poopflinging of two sides sucking their brand creators dicks, but it would be much better if there was more, imagine if Voodoo was still here, they'd problably have a nickname of Doodoo and GPU threads on /g/ wouldn't exist because all they would be is shitposting.

>>57736807
Pretty sure AMDs marketing team isnt made up of fan boys retard.

The execution engine is nearly identical to haswell. it has a few extra FPU ports. its L1 instruction cache is bigger. its L3 cache is a lazy victim cache and will probably be shittier. communication between the multiple L3 cache modules will be slow compared to intel.

otherwise there are very few architectural differences. the blender demo was more than likely a fair demonstration of its general IPC.

the big unknown is still clock speeds and i'm not optimistic given how dogshit 14LPP is. However Broadwell-E also clocked like horseshit so the difference may not be all that massive afterall. It'll probably end up a good buy, that is until Intel finally releases Skylake-E, which will destroy it.

>>57737093
>probably true

lol

>>57757394
amd released everything form the demo, it was gone though, and blender was not fucked with if memory serves.

>>57757506
its a stressful benchmark, that and there are renderers that do not use the gpu

>>57762647
Manic Indians.
Even worse.

>>57762841
>until Intel finally releases Skylake-E, which will destroy it.
AMD already have new tech process planned for future, too. Lets hope it wont be A10 shitty APU's analogue.

>>57757506
Bulldozer was good back then.

File: 895bfdcf6905c3df19e3d1695feba33f.png (2KB, 280x58px) Image search: [Google]

2KB, 280x58px

>>57740551

>>57763215
Yeah the 7nm. Which is actually closer to 10nm. And will probably be late and shit, knowing GloFo.

There might be some decent optimizations of the 14nm (or 20nm rather, since that's what 14LPP really is) by the time Zen+ is released, but by then Intel will be on Cannonlake, in fact they might even have a brand new architecture.

>>57745217

I was talking about F5 Networks, which I founded with a friend of mine.

>>57763374
confirm on your company twitter pls

>>57763296
if you combine an AMD cpu and a NVIDIA GPU, would it be possible to emulate the power of a sun and harness free power?

File: 5ce4828a3149ae30828ed9fb53e5b466.png (18KB, 563x344px) Image search: [Google]

18KB, 563x344px

>>57763508
you tell me

>>57763545
>24 C GTX-1060
>69 C A8-5500
something seems wrong here, unless you for some reason applied a CPU heatsink to the GPU or something, because that is just nuts

>>57763741
Speccy can't read AMD CPU temps right my newboi

File: Piriform_Speccy_2016-11-30_08-04-07.png (2KB, 271x62px) Image search: [Google]

2KB, 271x62px

>>57763753
maybe newer CPUs

>>57763741
I thought you were going to ask how someone stuck with an A8 can afford a 1060.

>>57748307
Compile chromium

>>57736807
>trusting sleazy media
i bet you voted shillary too.

>>57759421
go to /pcbg/ and post build with AMD see them swarm like fish to bait

>>57765569
I bet you read fake news you putinbot

>>57740921
the amount of heat generated is the same you retard.
heatsink is for removing the heat from the cpu to the enviroment and the h80 do it faster.

>>57763545
>2048MB NVIDIA [...] 6GB [...]

..what am I missing here?

>>57762617

were already approaching the limits of micro architecture

>>57736807
So that makes it about as fast clock for clock, core for core as Sandy Bridge.

>>57765823
4096MB

File: U4fVVHE[1].png (51KB, 836x868px) Image search: [Google]

51KB, 836x868px

>>57740921
>>H80i is designed to cool 140W+ CPUs.

Is that supposed to be impressive?

>>57745420
Let me shoot you in the foot real quick to improve your walking performance after 3 months of recovery.

>>57765641
What "fake news" do you speak of? Please enlighten

File: 1477807332791.jpg (302KB, 1454x993px) Image search: [Google]

302KB, 1454x993px

>>57766740

>>57766740
after trump victory google, Obama, acebook and every Mainstream news went into attack against alternative media and declared they share "fake news" and facebook began wiping them from FB, google took adsense money and obama wants to new laws

>>57756020
>>57756305
being close to the E5-2630v4 would be pretty nice if it overclocks ok

gaymers might not like it but that'd be solid for creative work

if it delivers a lot of pcie lanes on the cheap it may well find love among the gpu-heavy crowd like deep learning and video production

>>57741086
Can tell you that the xfx gtr (assume the 470 go is the same if not better) runs cool and relatively quiet so long as it has plenty of air being fed to it. Dont really hear it till you break 65% fan speed.

>>57768816
Im a bit worried AMD may jew us on the pcie lanes on the $350 sr7. What could command a $150 difference between the two sr7's?

>>57769078
This
Also the whole clocked to 3ghz 6900k vs zen blender thing was a way to push prices imo.

>>57769078
Yeah I hope they don't cuck the lower models too much, a full set of pcie lanes could find them a lot of fans and finding a niche like that might be more valuable than just forcing everyone to buy the expensive part/continue buying intel

I wonder what goes into that $150 price difference, could just be the top binning with best clocks and a shitty yield though? Guess we'll find out in a couple weeks

>>57769078
I wouldn't worry too much, few people really "need" all those lanes, and the ones that do wouldn't mind paying extra when you compare prices vs Intel's 40 lane offerings.

>>57769390
10 WEEKS LEFT

>>57769606
They're doing some demo event shit in like two weeks (dec 13), hopefully we learn all we need then.

Because if it sucks I'll be pulling the trigger on a xeon workstation build to replace my old 3930K work pc.

>>57737615
>anally raped at every price point in the GPU market by Nvidia

Not really. The 280x and the 380 are much better and cheaper than the 1050ti.

And the 480 is cheaper than the 1060 with similar results in most games.

Low and mid end is all amd. Nvidia is for higher end.

>>57770409
you wish. thanks to marketing and reputation even nvidia is going after those categories not sure if really paid or not but on most build-a-pc they recommend nvidias the most

>>57770340
After that Blender demo, I have a feeling that we will just have more questions than answers until 3rd party reviews and shit come out. We already know its "capable" but how much so?

>>57770493
Well the right choice is amd for low budget. Better dx12/vulkan support and so on.

>>57736807
NOT EVEN FASTER THAN SANDY BRIDGE HAHAHAHA

>>57770638
False

>>57770409
Maybe if you buy used. Personally I'd say the best used deal is the 380x. It's selling for about the same price as the 380. 280x is too hot

>>57770701
Prove it faggot
There is substantial evidence to suggest that Summit Ridge isn't twice as powerful as Piledriver, doesn't have the same single-threaded performance of Haswell (or even Ivy Bridge), and can't reach high frequencies for shit. The SR7 version can't even compete with a six-core Broadwell-E, so they're pricing it under the i7-6850K because they know it's going to be shit.
D
O
AMD

>>57770615
truth barely matters nowadays

>>57770998
Read the thread

>>57771014

/g/ is violently allergic to the truth.

>>57744354
>1997
>Buying anything but revolutionary Matrox
>3dfx cucks BTFO
>3D
Get with the times, 3dfx might have built the launchpad, but we MATROX boys gonna take you to the stars.

>>57745217
Yeah but SA claims every other day that AMD will go bankrupt by next week.

>>57740182
>macedonian detected

>>57771484

SA is a blatant attempt at stock manipulation.

>>57736807
>Jacques Faucault
>Trusting french fries

File: nervous.jpg (64KB, 474x350px) Image search: [Google]

64KB, 474x350px

>>57759901
Tfw I bought a 6700k for the single core performance rather than the extra threads

When is Zen out?
Would I want Zen if I do video encoding and 3d rendering?

>>57772025

>muh L3