Icekake confirmed successor to 8th gen and confirmed 10nm+

tbhq, for laptops it will be pretty good.

> inb4 edgy Ryzen shills
I know Ryzen is fucking amazing for desktop, but Intel makes amazing laptop processors and AMD won't compete until we see Raven Ridge benchmarks.

>>61909252
>amazing laptop processors
what did he mean by this

Coffeelake isn't even out yet and Intel's 10nm is an abortion of epyc proportions.


tl;dr Icelake is skylake with 3%+ more IPC and it will be launching in the same timeframe as Zen2, I wish it best of luck because it will fucking need it.

>>61909321
Coffee Lake ≠ Icelake

Icelake will focus on mobile.

You seem angry and confused, pajeet.

This is a white/jewish thread, not indian/chink.

Get out.

File: justfuckmywife8-min.jpg (43KB, 1002x1480px) Image search: [Google]

43KB, 1002x1480px

>>61909366
But Brian you told us Cannonlake focuses on mobile?

>>61909233
>literally copying a reddit title including typo

>>61909366
Cannonlake is mobile focused you fucking favela monkey.
Icelake is the next "architecture" aka tock like Skylake was.

>>61909391
Should've been Icekike

Just leaving this here.

>>61906191
>>61906205
>>61906218

>>61909491
Everyone knows that Intel's 10nm node is a disaster.
It's what, two years late?
Fab leadership my ass.

>>61909288
Intel's low power and laptop processors are pretty good

>>61909607
I hope you understand that laptop processors are simply desktop cores clocked lower with most of the uncore cut out?

>>61909395
>t. Rakeesh Zakari

>>61909321
Zen2:
> No multi-core advantage (like desktop Ryzen vs Desktop Intel)
> Even lower clocks due to mobile limitations (rumors stating a max of 3.3GHz on boost)
> Extremely cut-down Vega graphics probably won't even beat Intel Iris Plus (No HBM2, only 700 processing units and extreme thermal limitation)

http://wccftech.com/amd-ryzen-4-core-8-thread-raven-ridge-benchmarks/

Ryzen shines on servers and desktop, it will have mediocre performance on mobile, and will only compete with dual-core i5's and i7's.

Nothing will beat those 45W quad-core i7's with Iris Plus graphics for a while.

>>61909670
>currytech
>here is how a chip that isnt out yet will perform

>>61909670
HAHAHAHAHHAHAAHAHA
NOW THATS WHAT I CALL DENIAL

>>61909233
>How can AMD even compete?
With 7nm of course.

>>61909670
Raven Ridge has nothing to do with Zen2 you troglodyte.
Get a clue.

There won't even be 45W mobile Raven Ridges either, only up to 30W because 45W nonsense is paired with a discrete GPU and goes into alienware-tier junk

>>61909670
Raven Ridge is not Zen2, retard

>>61909692
Sorry, I meant Raven Ridge

>Cannon Lake
>Coffee Lake
>Ice Lake

What are all these Lakes and which one is which and which one even fucking exists? What the hell Intel?

>>61909713
Your premise is completely wrong regardless, what determines Raven Ridge's success is how it does at 15W.

>>61909730
And it will do well.

>>61909670
>No multi-core advantage (like desktop Ryzen vs Desktop Intel)
You realize most of the Intel mobile crap are overpriced dual cores, right?
> Even lower clocks due to mobile limitations (rumors stating a max of 3.3GHz on boost)
That is pretty damn good for mobile.
> Extremely cut-down Vega graphics probably won't even beat Intel Iris Plus
Intel Iris Plus is their absolute top-end shit that goes into $400+ laptop CPUs.

>>61909722
*lake - designed by israel team
*well - designed by USA team

>>61909730
I hope for a 4-6W fanless SKU personally.

>>61909670
You do realize that single-core performance is very close to Kaby Lake with a much better efficiency right? Those 4C/8T Ryzens will be competing with mobile i7's with 4c/8t which are expensive as fuck. Vega could be good efficiency at lower voltages/clocks for all we know. AMD loves to overvolt the shit out of their chips.

Raven Ridge will destroy Intel CPU's in performance and power.

>>61909730
Zen's 14nm is literally designed to undervolt and run on lower power, that's why it has trouble getting above 4GHz.

>>61909773
Don't lecture me I know that, AMD's job is getting uncore power draw down because uncore is pulling 20W alone on Ryzen 7, around 100W on a EPYC.

>>61909801
Well RR will have LESS uncore.
Zeppelin has a lot of it.

>>61909321
>abortion of epyc proportions
simply epic

>>61909801
>uncore
The what now?

>>61909801
Then that means the cores are just ultra efficient, making an efficient uncore is easier (it's just a bunch of SerDes) than efficient cores, AMD is on the right track

>>61909841
/g/ - technology

>>61909744
>You realize most of the Intel mobile crap are overpriced dual cores, right?
Most of it? Yes. But not all. Most of workstation laptops have quad-core i7's
> That is pretty damn good for mobile.
True, but nowadays Intel mobile i7's are getting 4GHz, Raven Ridge can't compete against that, not without the core advantage.
> Intel Iris Plus is their absolute top-end shit that goes into $400+ laptop CPUs.
Yeah, but Intel HD 620 for instance is still a pretty decent and popular iGPU, Raven Ridge will probably be close to that on graphics.

The problem with RR is graphics. Ryzen was a success but Vega was a massive failure, I bet my ass it won't scale down so well like those AMD slides, and that's the reason we still haven't seem any preview from AMD.

My bet is: It will beat dual-core i7's by a little margin, having only 5% better graphics.

Pretty good for the average consumer, but not enough for workstations.

>>61909858
>marketing buzzwords are technology

>>61909877
Jargon you don't understand are not buzzword you stupid monkey.

>>61909873
You can literally slap a 1700 in ebin werkstation laptops.
And no Vega is still ages ahead of Intel iGPUs, despite being a bottleneckfest.

>>61909873
What? Even underclocked GCN2 is a match for whatever Intel has, it only came on top when Intel used more silicon for the graphics than the entire AMD die + the advantage of 14nm and actual good cores.
CPU side will be close, but the GPU will absolutely demolish Intel.

>>61909919
>Vega is still ages ahead of Intel iGPUs, despite being a bottleneckfest
[citation required]

>>61909873
>True, but nowadays Intel mobile i7's are getting 4GHz
Link one that doesn't throttle.

>>61909883
Uncore is LITERALLY something that Intel uses. In reality it's a northbridge inside the CPU(SoC design).

>>61909935
>GPU will absolutely demolish Intel
Vega is very problematic tho, without HBM it may even be a downgrade on performance.

>>61909958
I see you're absolutely clueless about this subject.

>>61909958
>very problematic
Why.
It's 700 something ALUs.

>>61909950
>Link one that doesn't throttle.
So what? Throttling isn't always bad, short boosts are very effective. AMD does literally the same with XFC.

>>61909974
Vega64 is literally slower than Fury X on the same clock speeds.

What clock speeds do you expect to get on RR GPU? Now remove HBM2.

>>61909982
>moving goalposts: the post

File: 1497184864971.png (25KB, 554x166px) Image search: [Google]

25KB, 554x166px

>>61909670
LMAO

>>61910002
Nope. Are you implying they will throttle down to speeds lower than 3.3GHz?

Also, are you implying Raven Ridge won't throttle as well?

>>61909997
So is Pascal compared to maxwell, doesn't stop it from going into mobile.

Clocks don't matter, efficiency is all that matters in mobile.
And efficiency is usually gained from low clocks.

>>61910026
> Ryzen at 5GHz
That kills Intel.

>>61909745
*ganges - designated by India team

File: 1502151449208.png (55KB, 1509x905px) Image search: [Google]

55KB, 1509x905px

>>61909730
>>61909801

>>61910037
> Vega
> Efficiency

Ryzen is very efficient, but Vega? Hell no.

File: 1484031755829.png (156KB, 564x926px) Image search: [Google]

156KB, 564x926px

>>61910045
Based IBM

>>61910080
You do understand that Vega10 is 4k ALUs?

>>61910029
7700HQ base clock is 2.8Ghz for a reason.

I'm saying that your claim of 4Ghz is bullshit. I'd be surprised if any laptop can get even 3.5Ghz out of Kaby Lake without throttling down after a few seconds.

>>61910080
Efficiency isn't a linear scale. We don't know how it will perform with low voltages and clocks and less CU's.

>>61910080
So you are retarded, they're not sticking a 4096 shader part with 1700MHz clocks into a fucking 15W APU you dumb shitface, it's a 11CU part with sub 1000MHz clocks, for fucks sakes there's already a 150W Vega Nano in the works and that's still 4096 shaders.


Get a clue, stop embarrassing yourself.

>>61910089
Why doesn't Intel just drop the 10nm+ bullshit and just use that shit as well? Everyone else will be using that (Apple, Samsung, Qualcomm, AMD, etc).

>>61910121
>why doesn't intel just drop their core business

>>61910121
>why doesn't Intel use someone else' fabs for its Core arch

Gee I wonder.

>>61910111
What makes you think a cut down Vega will be more efficient than Intel Graphics? There's literally no source on that.

The only thing we know is that Vega was a disaster on desktop. With that is possible to assume that Vega will underperform on mobile as well.

> inb4 just wait

>>61910138
> AMD Ryzen is IBM Ryzen

>>61910161
Because Intel graphics are fucking garbage and there's no Nvidia there.
And because every other GCN version undervolted fine and ran efficient at those power targets.

As said, there's already a 150W Nano in the works that will be some 10-15% slower than a 300W Vega, it's not rocket science, voltage/power curve is not linear.

>>61910185
>I don't know that Intel literally competes with other fabs BECAUSE THEY HAVE THEIROWN FABS
It's literally how they've been ahead all this time you fucking moron. Shut the fuck up if you don't know anything.

>>61910121
Intel worked really hard to get into this situation in the last several years.

>>61910161
Let me make this simple for you.

Even if AMD used GCN1 in Raven Ridge, it would still have better graphics than Intel.
Intel GPUs are such pieces of shit that they have trouble running Diablo 2 without stutters(obvious lack of legacy 3D driver hacks from the GMA days)

>>61910161
>possible to assume that Vega will underperform on mobile as well.
>possible
>assume

Even doing that it will still be better than intel dogshit integrated graphics

>>61910252
Maybe D2 had unoptimized code. HD530 can run D3 in max settings at 1080p 60fps.

>>61909745
>the last real good Intel CPU was Broadwell and Haswell

Hmmm? Really makes you think

I bet Icewell would shit on AMD so bad the courts will drop them with anti-trust and anti-monopoly shit but gotta give shekels to Israel

>>61910334
Broadwell was a fucking disaster, Haswell was great.

>>61910334
It's not Icewell, it's Icelake , meaning it's another Israel design

Intelfags sure have some wild dreams, too bad they're usually fueled by their lack of knowledge.

>>61910334
If current Intel can't solve the huge dies issues and the thermal problems then a theoretical Intel wouldn't be able to either. It's not Intel that's bad, it's that ryzen too good and they had to rush into territory they didn't explore.

>>61910362
reading comprehension

>>61909640
delet.

>>61910339
Actually it wasn't, 14nm was a disaster and so has Skylake and beyond been a distaster.

After 14nm being so late delayed Broadwell a few times, it was found to have BETTER performance than Skylake and was shelved for desktop.
Compare Broadwell-E to Skylake-X performance, Broadwell-E kills it, every time.

Intel is actually getting WORSE over time.

It seems like Raven Ridge will be using either Vega 10 or Vega 8 graphics, so 11CU and 8CU(what previous iGPU used)

Also it seems raven ridge isn't a 1XXX part, but a 2XXX denoting a new generation, when it launches it'll be closely followed by Pinnacle Ridge.

>>61910456
Intel made a tradeoff with Skylake-X, mesh(well, a mesh of ring buses) over ring bus for higher 16+ core throughput, and L2 over L3 for AVX512 and large data performance.

Skylake-X isn't worse, it's just different, and that difference negatively affect gaming performance as gaming likes low latency cores, memory and shared L3 cache.

>>61910456

>tick
>tock
>just

>>61910483
Zen is no different, it trades latency and shared caches for modularity as workloads that actually need 20+ cores and are latency sensitive are rare, latency isn't scalable like throughput.

Personally I'm glad both Intel and AMD give gamers the middle finger, who the fuck do they think these companies should center their architecture around their shitty needs?

>>61910483
It's worse in every metric, every benchmark...
It's even more expensive...

Skylake-X is shit for everyone.

>>61910562
Zen is still the future though, you can't go much forward with clock speed on few cores. Intel held back technology because of the easier production which led developers coding for single core performance. When 8+ cores becomes standard they will learn to optimize games for multicore and then you don't need those things you mentioned.

>>61909653
SOPA

File: 67413maxjops.png (33KB, 650x525px) Image search: [Google]

33KB, 650x525px

>>61910625
It's not, stop shitposting.

>>61910700
>22c vs 28c
Wow more cores for a parallel workload = more score, who fucking knew!

>>61910774
You're conveniently ignoring that these workloads have been tailored for Intel's previous cache hierarchy of 256KB L2 and 2MB L3 per core for the last 10 years, both skylake-X and Zen need optimizations and they'll get it, the gap will only widen, it's already ahead at this early stage.

You're not smart, just stop.

>>61910700
>>61910774
Also I should point out that

53,052 / 22 = 2411.45 average per core score
60,693 / 28 = 2167.61 average per core score

Both CPUs have 2.8ghz max all-core turbo.

>>61910805
The scores are worse, period.
>bbut it might be better some time in the indeterminate future!!!
I think AMD has a GPU to sell you...

>>61910847
Scaling isn't linear you idiot, IPC is a dickwaving metric on forums, comes only into play after efficiency, features, memory and performance, if it's faster at the same power, you achieved your goal.

And it's a fact that server workloads are tailed to uarches, don't be retarded, this isn't the gaymen market where we still use MMX and SSE1

>>61910886
>it's a fact that server workloads are tailed to uarches
in HPC maybe... clearly you've never worked enterprise.
Moreover, considering that statement - Skylake-X will be handily defeated by Epyc anyway - so it's still a shit arch.

>>61910940
Intel didn't expect Zen to be that good, that's their oversight, they were trying to outpace Broadwell-EP, and they'll suffer for it.

To be fair AMD purposely sandbagged, if the IPC was 15% lower(intial estimates) Intel wouldn't have much to worry about.

File: 1500824087430.jpg (9KB, 284x177px) Image search: [Google]

9KB, 284x177px

>>61910225
I hope this buries them

>>61910562
>Personally I'm glad both Intel and AMD give gamers the middle finger
They're giving lazy devs the finger. A properly threaded game doesn't have significant issues with multithreading. Oh noes they have to think about engine architecture now

>>61911006
It takes two to tango, enterprise can pay for good coders that can get around the mutex horror shows of C++, game studios don't care, they're left waiting for better tools like c++ 20 and 17.
Rust is too young still, it has a very good multithreaded design, but still too young.

>>61910531
underrated post

WAAAAAAAAAAAHHHHHHHH
INTEL ON SUICIDE WATCH
KRZANICH LITERALLY JUST JUMPED OUT OF A WINDOW
and he blamed trump on the way down, but we all know the truth here.

File: comfyAmada.jpg (10KB, 225x225px) Image search: [Google]

10KB, 225x225px

>>61911090
AMADA is at comfymus-maximus levels

File: 1499957247669.png (84KB, 653x726px) Image search: [Google]

84KB, 653x726px

>>61909670
hehe good old reliable wccftech setting the record straight as usual

>>61911006
this
fucking devs had issues with jaguar because MUH SERVAL PROCESSORS
then they had issues with Cell because MUH 8 CORES
now they have issues with zen because MUH NO 10 GHZ PENTIUM-X
learn to code monkeys or back to the slums

>>61911090
oh shit, I really should read a tad more slowly.
he just jumped out of some stupid council window, it' wa probably on the first floor.

>>61909233
>Icekake

Is there any way we would see 6c/12t and 8c/16t in laptops chips from AMD?

>>61911277
6C APU? only with zen2

>successor to
it's dead, jim

>>61909233
The last well made processors of Intel were sandy bridge iX 2XXX. Then they relaxed and inertially collected shekels. Intel could even just stop producing any kind of processors at all and they still would earn 10 times more shekels than AMD

>>61911572
They would earn next to nothing with no CPUs.

>>61909233
With 5 nm GAAFET

>>61909233
>inb4 icekike

>>61910700
That's not Skylake-X retard.

>>61912396
8176 is SKL-X.

>>61909745
What about Sandy Bridge?

>>61912583
Haifa.
It was initially called Gesher afaik.

>>61909607
Yeah, if you love dual cores. In 2017.

>>61909745
So, stagnation is actually jews fault?

>>61912676
it's a meme for a reason

>>61909640
Top tier processors from one sillicon die are for servers,
tier below that goes to laptop manufacturers,
and bottom tier are desktop processors.

It'll be competing with Zen2 at 7nm.
Good fucking luck

How long will i have to wait till i can get 6-8 cores that run at 4-4.5 Ghz for ~300-400$ i would like to upgrade my 4790k one day

>>61913120
Next week when Covfefe Lake launches. You're better off waiting a year for Zen 2 though.

>>61913177
isn't coffee lake at around 3.5Ghz?

>>61913326
4.3GHz six core turbo out of the box.

>>61913326
Who knows? Wait until it comes out before making judgements.

>>61913358
oh i see, i think i should have been more clear then.
I was referring to a 4 Ghz base-clock with around 4.5 Ghz turbo
>>61913377
that of course is the only reasonable thing to do.
i was just curious what the expected future might be

>>61910121
Because they probably make up to 70% profit out of each chipset sold just because they have their own fabs and don't have to pay someone else to do it.

>>61913403
You're just retarded then. Base clock is utterly irrelevant as far as Intel chips go and so is 200MHz. If you want more, overclock.

>>61912597
>mfw i7-6700 low power version

File: 1472522394684.jpg (185KB, 1360x1309px) Image search: [Google]

185KB, 1360x1309px

>>61913452
ok call me once the boost clock gets to 4.5 retard
the boost clock will only reach 4.3 on a single core
assuming this pic is true
the boost for 4+ cores is 4Ghz
i want to have 4 cores that are able to reach 4.5 Ghz
which none of them will
and an overlock that would enable such a thing would most likely lead to a housefire

>>61910089
Jesus, I knew about the higher clocks, but lower power and die cost too? Next I'm gonna find out the yields are going to start kissing the centennial mark.

>>61909233
>Ice Lake
>Ice
That means solder, <65W TDP, stock cooler keeping it under 50 degrees running Prime95 small FFTs at 5.2GHz incoming.
AyyMD is finished.

>>61909233
>How can AMD even compete?
simple amd will release an actual product that gives gains over the previous generation

>>61909801
hence why pussy keller came to work on the mobile chips and k12..

File: 1501584582323.png (418KB, 627x627px) Image search: [Google]

418KB, 627x627px

>>61913929

>>61909607
>>61912597
Stop living in the past, you could get a 4770K in laptops for years now, peasants.

> inb4 waaah it costs moneeey
Fuck off, neckbeard.

>>61909252
Wanted to buy a 7xxx series i7 laptop to replace my 4700HQ, but will wait for Ice Lake.
4700 -> 7000 series already had like ~25-30% improvement.

>>61913952
> AMD is shit when it comes to IPC
> new CPU by Keller
> it's shit when it comes to IPC
eh.

File: 1502488836418.jpg (24KB, 267x297px) Image search: [Google]

24KB, 267x297px

>>61909491
DELID DIS

>>61914957
Thank you. You might want to also spam /b/ and /r9k/ threads, just so your "work" does not go to waste.

File: 1489122339460.png (495KB, 1070x601px) Image search: [Google]

495KB, 1070x601px

>>61910531

File: sketch1502738459632.png (519KB, 952x644px) Image search: [Google]

519KB, 952x644px

ANOTHER Skylake CPU, but this time on 10nm
Wow, amazing

>>61914980
ok schlomo

>>61915000
>>61915025
>>61914957
You are really getting worked up here, anon. Remember what the doctor said, it's bad for your heart.

>>61909233
>10 nm
Too little too late. Everyone wants to see 7nm.

>>61911006
This
Competent devs have no problems with the 1.6GHz Jaguar 8 core in the PS4

File: 1502378315806.gif (127KB, 600x600px) Image search: [Google]

127KB, 600x600px

>>61915041
I can't take this anymore

>Throttling isn't always bad
/g/ - goynology

>>61909252
>throttling pile of garbage
>""""""""amazing laptop processors""""""""

File: 7 Core comparison to Jaguar.png (359KB, 1914x1076px) Image search: [Google]

359KB, 1914x1076px

>>61915059
>Jaguar 8 core

AMD has a strange way of counting to 4.

>>61915418
>Throttling doesn't matter!

>>61915752
>brazilian education

>>61915096
> implying people care about these "news"
Nigga if I have to replace my PC, I have to replace it.
At the moment there is only Intel in mobile space, Ryzen simply sucks. On desktop it's the same thing, single thread wise Intel wins all the time. So unless I want to build an LHC in our office building, why the fuck would I consider buying a Ryzen CPU at all?
There is really no competition, it's a different market.

Heavy workstation? ThreadRipper.
Anything else? Intel.

>>61915418
> throttling
LMAO stop buying laptops from eBay.
My 4700HQ runs without any throttle overclocked in my shitty plastic ASUS ROG laptop.

> muh grapes
Fucking faggot.

>>61909491
>semiwiki.com
people on /g/ are retards

>>61909252
>I know Ryzen is fucking amazing for desktop
You can undervolt it like fuck and it'll still run at like 3GHz.

>>61916602
> like fuck
It's silicon lottery. Some chips can be undervolted, some cannot. It's not a thing you should base your purchase on. If it were undervolt-able much, it would be present in laptops already.

>>61916569
All I see is AMD shilling and shitposting. That is very objective and so informative.

>>61916705
>If it were undervolt-able much, it would be present in laptops already.
Asus showed off a few already. You can't just shove a desktop CPU in a laptop with no R&D though.

>>61909491
In term of tecnology, ice lake is pretty good processor.
The usage power is lower while having better performance than 1st generation 14nm processor.

>>61916763
To be honest single thread wise it's shit, so maybe Zen 2 or whatever the next unit will be called.
Until then I only see the next-gen APUs getting a hold of low-mid price range market.

>>61916781
>To be honest single thread wise it's shit
Good thing all laptop processors have shit single threaded performance.
>b-but they don't
Literally check laptop users bitching in any emulator forum.
Also, single threaded performance will prob be shit until DDR5 since the infinity fabric shit scales with RAM clocks.

>>61911070
Rust will never catch up in games. Most engines are decades+ of investment by hundreds of programmers and alnost all of it is in c++. And so is all of the conventional wisdom and expertise.

>>61916060
>Multi-core doesn't matter!
>Price/performance doesn't matter!
>Performance per watt doesn't matter!
>Power usage doesn't matter!
>Temperatures don't matter!
>Soldered dies don't matter!
>Upgradeability doesn't matter!
>Anti-competitive business practices don't matter!
>*NEW* Locked platform features don't matter!
>*NEW* Synthetic loads don't matter!

>>61914924
>literally a wider core than skylake
>worse IPC
What are you smoking you dope?

>>61916812
https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-7700HQ+%40+2.80GHz
> 2025 SC
> shit performance

https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+7+1800X

Uhm, for sure!

>>61916845
I am simply saying I can sacrifice 2-4 cores for better single thread performance in a mobile form factor.

Reminder that Ryzen on desktop doesn't use the on-die voltage regulation. That's saved for mobile

>>61916893
>pissmark
lol

>>61916893
>i7-7700HQ
>Recommended Customer Price $378

Costs more than a fucking desktop i7-7700K, that's why. Meanwhile Raven Ridge will hit desktop clockspeeds at a much lower price.

File: india inside.jpg (387KB, 1500x1164px) Image search: [Google]

387KB, 1500x1164px

>>61917216
>bringing up completely unrelated gaymur shit
>POOLARITY
>Rajeesh

>>61917225
How does it even have a price when it's a mobile CPU?
And how the fuck can it cost so much, when a ROG costs what.. 1500$ with a 1070 or 1080?
I doubt ASUS would be selling these with no profit margin.

>>61917241
Because that is the only thing AMD posts and benchmarks all the time. And Tomb Raider maybe?
CPU wise they fap to Cinebench.

>>61917282
>How does it even have a price when it's a mobile CPU?
It's on Intel's website. https://ark.intel.com/products/97185/Intel-Core-i7-7700HQ-Processor-6M-Cache-up-to-3_80-GHz

>CPU wise they fap to Cinebench.
Cinebench is a lot more of a neutral benchmark than pissmark, which is absurdly Intel biased. There's at least 15% bias for Intel CPUs on there. http://www.cpubenchmark.net/compare.php?cmp[]=2966&cmp[]=2794

File: 9oVGc83.png (64KB, 960x540px) Image search: [Google]

64KB, 960x540px

AHAHHAHAHAHHAHAHAHAHAHHAHA
>this deep in denial
Zen will fucking destroy anything Intel can put out for the next 2 years in the mobile market.

File: 1497883390389.jpg (18KB, 267x297px) Image search: [Google]

18KB, 267x297px

>>61917387
NOOO INTEL WAS SUPPOSED TO BE BETTER

>>61909233
> still no 8 cores 16 threads for 350€

Shit. AMD is the only solution

>>61917811
What do you do on your computer?

> inb4 gentoo
Ryzen / ThreadRipper causes segfaults on BSD/Linux.

>>61917887
>What do you do on your computer?
It has nothing to do what I do, It's evolution. We've been stuck with 4 cores for over a decade now for this price. AMD finally offers 8 cores and 16 threads for that price but intel does not. Give me one single reason, besides shitty segfaults for systems nobody uses, to buy an intel

>>61917887
http://phoronix.com/scan.php?page=news_item&px=Ryzen-Segv-Response

>AMD was also able to confirm this issue is not present with AMD Epyc or AMD ThreadRipper processors, but isolated to these early Ryzen processors under Linux.

>>61917918
I am no shareholder at either.
Buy Intel if you Photoshop, do office work, develop, debug programs, virtualize, browse the web, use web tech, use script languages, emulate or maybe play games.
Buy AMD if you Cinebench or render all day.

Bretty simple. More cores mean jackshit if they just wait. And I don't mean to shill any brand above, rendering is indeed a legit task.

>>61917887
>Ryzen / ThreadRipper causes segfaults on BSD/Linux.
ThreadRipper and Epycs cpus are not effected.

>>61917957
Correction: If you virtualize large clusters, obviously AMD is a better choice there. But I don't think many desktop computers would ever do that. That's already a server task.

>>61917957
>develop, virtualize
These are better on Ryzen since more cores = faster compile times and better for VMs.

>do office work, browse the web, use web tech, use script languages
This is all shit a low-end potato PC can do anyway.

>emulate
The only emulator that requires high clockspeed Intel is CEmu and that is rapidly changing as it gets optimized.

>>61918000
> develop
Not sure how many projects you wrote, but most source trees don't have a parallel build system. I mean if you write a program for yourself or a small business, you won't bother with it. Debug too, runs on a simple core. The IDE also with it's plugins eat a single thread.

Indeed you can use several cores on Gentoo. Very useful indeed.

> inb4 but companies
Companies have buildbots. No one will build the software on their machine. You submit your shit to git/svn or whatever and the bots will build it for you.

>>61912011
More like FAGET

>>61912597
I love to be DP'ed
t. Intel fanboy

>>61918042
You still need to build local if you debug with a ide like visual studio or eclipse.

>>61918465
Then the debug process is single threaded. Back to square one to be honest.

>>61918465
>Not using remote debugging, even gdb
>Using eclipse and caring about performance
Did you finance the rims on your 04 Civic?

>>61915046
Isnt Intel 10nm sort of equivalent to 7nm? Like tsmc 16 nm and everyone else 14nm

>>61920868
Intel themselves say this.

Even if it's true; Zen2 on the equivilent of intel's 10nm? Intels' fab tech is the only thing keeping ryzen from turning their busted anuses into prolapsed ones

>>61917887
>straight up lying
The issues were resolved in Ryzen chips produced in May or June, long before epyc or threadripper were produced. Intel shills are unbelievable, though I guess this is just core Intel tactics

>>61920868
Yes, the competing 7nm node are about as dense, but every single one of them boasts much more advanced FEOL than Intel's 10abortion.

>>61920922
>The issues were resolved in Ryzen chips produced in May or June, long before epyc or threadripper were produced. Intel shills are unbelievable, though I guess this is just core Intel tactics
Jesus Fucking Christ, we only have AMD's 'promise' that Epyc or Threadnigger's aren't 'affected', when they themselves are not sure what causes the problem. We will see when Phoronix releases test results (because he is the only one testing for it).

And you are wrong: I have 1700x from launch day and another produced in May (both which are same stepping: there is no proof of b2 stepping yet) and both are segfaulting during multiple make operations on Taichi and C6H with latest UEFI (with 1.0.0.6 agesa), SMT on or off, doesn't matter.

>>61920902
Okay, and how glofo zen 14nm compares to Intel skylake/kabylake 14nm++ node?

>>61918042
>Companies have buildbots. No one will build the software on their machine.

Thanks for proving you've never actually worked in any professional environment. Not even the continuous integration meme calls for integrating every change you make so you can let servers build for you. In practice people will push to the repository at most once a day, often less. Most of the builds you'll do will always be on your machine

>>61920994
>and another produced in May (both which are same stepping:

>I just checked the date believe me!
>I knew about it all along and checked the date even though only recently people have been discerning date of production to pinpoint the problem!

>>61921169
I got the second 1700x last week. There is a week printed on the heatspreader (22). Nowhere did I wrote that I knew about date of chips onwards, where the problem supposedly should been fixed when I installed the chip last week. But only Sunday I changed heatsink and the chip from Taichi to Crosshair to check if segfaulting happens on it too and did.
It isn't fixed on week 22 1700x and that is fact.

If anyone is doing mulitple make operations (in other words emerging on your gentoo box), you won't be safe with 'fixed May or June' chips, because it isn't fixed. Phoronix will test the threadniggers and report if 1) multiple make operations or 2) kill-ryzen test will cause segfaulting. There is no reports yet, because none of the big sites test it.

>>61921304
And 'long before Threadniggers were produced', lol, because there is week 22 Threadniggers on circulation judging by multiple reviews.

>>61921304
Did you set soc voltage to 1.15v?

File: intelake.png (52KB, 795x495px) Image search: [Google]

52KB, 795x495px

>be intel
>have tick tock strategy
>have codenames represent this
e.g. sandy/ivy bridge, has/broadwell, sky/kaby lake
>fuck up scheduling after 14nm
>covfefe lake
>cannonlake
>ice lake
>tiger lake
>gemini lake
>crescent lake

why

>>61909942
>he really thinks Intel's shitty iGPUs come close to anything AMD or Nvidia have.

>>61921748
It was skylake and cannonlake, originally, with a complete new naming scheme after that.
Then the ice and tiger lakes were named and supposed to be the successors of cannonlake.
The problems started after that with kaby, coffee, etc

File: adskjfadskf.jpg (453KB, 2118x1184px) Image search: [Google]

453KB, 2118x1184px

>>61922988
I just figured out what Intel is using for TIM.

File: Intel exposed.jpg (804KB, 1846x1923px) Image search: [Google]

804KB, 1846x1923px

>>61912676
Who would've thought

>>61917387
Sweetspot is right at 35W though.

>>61925718
That's for 8cores.

File: intel 8.webm (3MB, 1920x910px) Image search: [Google]

3MB, 1920x910px

who else hyped?

>>61925778
This is some bibelines4-tier advertising.

File: 1500516507122.gif (61KB, 300x351px) Image search: [Google]

61KB, 300x351px

>>61909670
>a processor built upon a low power process and with actual graphics IPs will somehow lose to 10nm that's worse than 14nm+++++++ and a joke igpu that needs expensive eDRAM to even match last-gen APUs
nani

>>61925778
Is this an Apple ad?

>>61925767
Even better. A quad core will even do better in terms of power.

File: count.jpg (33KB, 660x396px) Image search: [Google]

33KB, 660x396px

>>61925778
fucking kek

>>61923854
Too expensive, cheaper to use Brian's jism

>>61926054
Brian isn't nearly fertile enough for that.

File: intel boss.jpg (123KB, 638x599px) Image search: [Google]

123KB, 638x599px

https://arstechnica.com/tech-policy/2017/08/intel-ceo-brian-krzanich-quits-trump-manufacturing-council/

S U I C I D E
U
I
C
I
D
E

File: the moneyshot.png (402KB, 617x463px) Image search: [Google]

402KB, 617x463px

>>61926157
That's what you think.