[ 3 / biz / cgl / ck / diy / fa / g / ic / jp / lit / sci / tg / vr / vt ] [ index / top / reports / report a bug ] [ 4plebs / archived.moe / rbt ]

Due to resource constraints, /g/ and /tg/ will no longer be archived or available. Other archivers continue to archive these boards.Become a Patron!

/g/ - Technology

View post   

[ Toggle deleted replies ]
File: 1.89 MB, 484x682, JUST.png [View same] [iqdb] [saucenao] [google] [report]
62540453 No.62540453 [Reply] [Original] [archived.moe] [rbt]


>> No.62540473
File: 1.04 MB, 400x224, FEELS GOOD.gif [View same] [iqdb] [saucenao] [google] [report]

>The 12nm switch allows AMD to roll out a new generation of RyZen processors as early as the first half of 2018

>> No.62540482
File: 246 KB, 882x758, 1489100922662.png [View same] [iqdb] [saucenao] [google] [report]


>> No.62540511
File: 387 KB, 1000x1000, JK.png [View same] [iqdb] [saucenao] [google] [report]

Now, everyone.
Let's say his name.
Say it LOUD and CLEAR.

>> No.62540531
File: 49 KB, 565x486, 1498759795881.jpg [View same] [iqdb] [saucenao] [google] [report]

>Literally not a single week without a massive fuck up on Intel's side since the very day Zen was launched

>> No.62540604
File: 162 KB, 633x900, ffffffvey.png [View same] [iqdb] [saucenao] [google] [report]

>> No.62540655

Holy fuck this is so JUST I can't believe it.
They should attach generator to Noyce' coffin for INFINITE ENERGY.

>> No.62540673
File: 66 KB, 785x554, 1505924291623.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62540681
File: 98 KB, 397x295, 14958027442730.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62540690

The only thing it destroys is Pajeet's last shreds of credibility

>> No.62540692


>> No.62540748
File: 2.88 MB, 415x232, 1502948246721.gif [View same] [iqdb] [saucenao] [google] [report]

But what will they do with all this TIM?

>> No.62540767
File: 1.23 MB, 712x604, YUNOBUY.png [View same] [iqdb] [saucenao] [google] [report]

>2000 mason tickets for 2 more cores and Turbo only on the two first cores, RF ID™ alongside with RAID KEYS™, DRYHORSE CUM™, INTEL DELID TECHNOLOGY™, LESSLANES™ and FUKUSHIMA POWAH & TEMPUSU™, oh and also don't forget to shekel out for 400+ dollarinos mobos
>Meanwhile 1950X is ~800$, has Turbo on all cores, and it's currently best mobo costs ~390$
Wow, what a hard decision on which to buy.

>> No.62540786

101% this

>> No.62540803
File: 3.26 MB, 480x358, 1505914570633.gif [View same] [iqdb] [saucenao] [google] [report]

Also no ECC RAM support with a $2000 cpu

>> No.62540816
File: 1.73 MB, 295x270, 1484247254566.gif [View same] [iqdb] [saucenao] [google] [report]

Oh shit, I completely forgot about that one, loooooooooooooooool!

>> No.62540822

it's just a refresh on new node, they won't even change the name, at least that's the only move that makes sense if they want to do tocktocktock like promised.

>> No.62540856
File: 345 KB, 800x600, 1502401815973.jpg [View same] [iqdb] [saucenao] [google] [report]

The raid key DLC nonsense has turned me away from Intel by itself. Jesus what a bunch of assholes

>> No.62540910
File: 1.49 MB, 2048x1536, 1504268873833.jpg [View same] [iqdb] [saucenao] [google] [report]

>> No.62541037

It's 2nm smaller/denser, you dumb fuck. -2nm allows fuch quite a bit more shit to be crammed in and power envelope with temps will be lower per clock. Meaning we'll highly likely to get RyZen 2xxx with stable 100% 4.2GHz on all cores & threads, and no more segfaulting since they've fixed it sometime ago. This will be THE perfect mainstream processor to get.

>> No.62541053

Who cares. No one *needs* a new processor.

>> No.62541096
File: 234 KB, 800x612, 1486816070397.png [View same] [iqdb] [saucenao] [google] [report]


>> No.62541134
File: 85 KB, 627x627, b16899965e048aff25784456ec792eddb1617e54d6c9eea281987487463a01fd.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62541162

Going by that logic no one needs a computer at all, so Intel can just stop existing at any time since we don't need them.

>> No.62541223

>Threadripper's currently best mobo costs ~390$
Which one is that?

>> No.62541243

Sounds like Taichi.

>> No.62541250
File: 618 KB, 1880x1455, Designare 001.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62541268
File: 377 KB, 1128x1376, Designare 002.jpg [View same] [iqdb] [saucenao] [google] [report]

Nah, it's actually seriously and absolutely unironically Gigabyte this time. They've literally managed to make a "perfect-on-all-fronts" mobo. It looks great, has all the needed features AND build like a fucking tank. This is THE board to get if you're wishing for best X399 experience out there, mateys.

>> No.62541381

dual socket x399 boards when?

>> No.62541404

you have to go EPYC to get dual CPU support, sadly.

>> No.62541406

Never. Buy EBYN.

>> No.62541425

mini ITX x399 boards when

>> No.62541435

Why would this be good news? It's obviously a stop-gap and both an excuse to delay, and an admission of 7nm Ryzen indeed being delayed.

>> No.62541439

It was always on the roadmaps you moron.

>> No.62541443

7nm hasn't been delayed you moron

>> No.62541448

Does it have a 10GbE NIC? And which one if true?

>> No.62541455

it's marketing-speak, you DUMB FUCK
the numbers haven't meant anything in literal ages

>> No.62541464

15 % density improvement is still 15% density improvement.

>> No.62541475

lol no it wasn't
it was just some vague BS about zen+ and zen2, then AMD implying + and 2 are one and the same 7nm update, with nobody actually knowing what was up

>> No.62541478

where the fuck did this guy learn to make great CPUs? is this a god-gifted talent? he just stops by a company, makes a great CPU architecture and then goes away to another, just to stop by again like an angel. he just magically touches a company and stocks suddenly increase.

how do i be like him?

>> No.62541481

7nm is totally gonna come out in 2018
you're totally not delusional

>> No.62541489

Only if you're dumb, roadmaps of Pinnacle Ridge were out a month after Zen1 was released, and that timeframe was way too early for Zen2 so Zen+ it is.

AMD also mentioned that they'd release new product each year (and they have to for shareholders and OEM to refresh)

>> No.62541492

Clark designed Ryzen though

>> No.62541504

>I-I knew it all along
you fags are just the worst

>> No.62541531
File: 134 KB, 1150x1266, sheseesyourpost.jpg [View same] [iqdb] [saucenao] [google] [report]

>I don't know what market momentum is: the post.

>> No.62541540
File: 127 KB, 1448x771, AMD-Pinnacle-Ridge-and-AMD-Raven-Ridge-APUs.png [View same] [iqdb] [saucenao] [google] [report]

>is angry other people know more than him about how business' and release cadence work
>he seriously thought AMD would sit on Ryzen 1XXX until 1H 2019

Literally retarded, must be Brian

>> No.62541552

The only thing I care about is whether or not the new node lets the chips operate with stability at higher clock frequencies. Current Ryzen was made on a node designed to run at 3GHz, but what about this one?

>> No.62541555

10% more performance, so 4.5ghz cap.

>> No.62541578

If we can get at least another 300MHz and a somewhere around 5% IPC, the threat Coffeelake will be much less.

It is a refresh on a slightly improved node, I'm not counting on miracles, I'll reserve that for Zen2

>> No.62541590

I really wanted to see what Zen would do on 12nm FDX(SOI)
The power would probably be nuts and design changes would also be insane but imagine IBM clocks on this thing.

>> No.62541598

They will simply do a round of price cuts to kill Coffin Lake.

>> No.62541599
File: 5 KB, 250x177, 1496252026564s.jpg [View same] [iqdb] [saucenao] [google] [report]

Somehow god keller in amd during ryzen, pure coincidence ?

Cpu design is team effort

>> No.62541600

It's slated for late 2018 or early 2019, retard

>> No.62541612

You mean 14HP?
12FDX is for RF and other low power shit.
Zen was designed by Mike Clark. Keller was doing K12.

>> No.62541626

keller designed infinity fabric though and ryzen wouldn't be what it is without that

>> No.62541640

The first thing to 7nm LP will be either RavenRidge2(EagleRidge?) or Rome(EPYC will solely need a refresh by 2019 because sapphire rapids ), client and threadripper will get last dibs.

So I don't think limited availability of 7nm RavenRidge2 APUs before Holiday season is impossible.

>> No.62541642

Keller is the head both arm and cpu

>> No.62541657

Stop shitposting.
Keller was doing K12.
Now THAT was a team effort. Including their semi-custom division.

>> No.62541699

They will probably refresh the entire product stack including EBYN and Vega10 on 12LP.
They need better EBYNs against Cascade Lake.

>> No.62541714

Stop shitposting Clark, you never be god like keller

>> No.62541717

Zen+ is a refresh of gen1 on 12nm, then Zen 2 on 7nm, then Zen2+ on 7nm, then Zen 3. Stop being such a retard.

>> No.62541729

>where the fuck did this guy learn to make great CPUs?
In his sleep while being in the womb of his mother, the Amaterasu?

>> No.62541733

Please stop shitposting.

>> No.62541734

Vega maybe, but EPYC isn't getting a 12nm refresh, only 7nm unless AMD thinks they won't have enough capacity on 7nm

>> No.62541742

>where the fuck did this guy learn to make great CPUs?
It's ArtX of CPUs.

>> No.62541750

head of both x86 and arm

>> No.62541752

If consumer Zen is getting a refresh than server one will get it too. Along with TR.
They all use the same die.

>> No.62541757

>where the fuck did this guy learn to make great CPUs?
During his sleep in the womb of his mother, the Amaterasu?

>> No.62541765

He was managing the teams. He himself has nothing to do with Zen. He designed K12.

>> No.62541769

fun fact: zen lead was a woman.

>> No.62541786

AMD's CEO is also a woman.
Who the fuck cares about gender as long as person is competent?

>> No.62541787

That's not in the roadmaps and a 10 month refresh on a server lineup is just a waste of resources, EPYC still isn't in channel sales because AMD can't make them fast enough for partners.

I can 130% assure you there's no EYPC refresh in 2018, but you might get Rome in 1H 2019

>> No.62541806

lel and neither do you if you're using it in this context

>> No.62541812

yup it just pure coincidence, I believe you

>> No.62541819

>he seriously thought AMD would sit on Ryzen 1XXX until 1H 2019
do you not read so good or do you have brain problems?

>> No.62541821

They will do EBYN refresh since Intel will refresh Purley in Cascade Lake.

>> No.62541833

it's literally time for krzanich to kill himself

>> No.62541839

No they won't, AMD won't release a new server lineup every 10 months, that's fucking retarded.

Purely has been selling since January, it's expected it gets a refresh, but the EPYC refresh which doesn't exist and 7nm Rome would be far too close to each other.

>> No.62541850

And unless Intel can get another 30% performance out of a refresh, then they STILL won't match EPYC, so AMD have no need to refresh, literally no one in their right mind would be buying Intel server processors right now.

>> No.62541851

They will refresh EBYN and TR simply because they are refreshing Ryzen. It's a byproduct of their manufacturing process.

>> No.62541868

Using the same die has literally nothing to do with it, lifetime, support and validation are the reason EPYC won't get a refresh.

Look at their fucking roadmap, what's next after Naples is Rome, which is 7nm.
No 14nm+(now 12nmLP)

>> No.62541884

>lifetime, support and validation
These are basically the same chips, no re-tapeouts required.
They can simply launch them as a newer stepping though.

>> No.62541891

question is it refresh or next generation
12nm doesn't sound all that exciting, it's barely change in density

>> No.62541894

It's a refresh.

>> No.62541898

no you can't, you have to do all the validation stages again

>> No.62541902

I dunno. Still seems like the same GoSlo LPP process, so clockspeeds won't be improved much.

>> No.62541904

It's a refresh, but it's not impossible for AMD to pick out some low hanging fruit out of Zen and increase IPC

>> No.62541924

considering memory controller was first generation and implemented long time before zen tapeout from another company they probably can add new one, it will increase performance quite a bit
10-20% clock probably can happen, 12nm LPP is somewhat like current 14nm from intel

>> No.62542004
File: 120 KB, 1920x1080, tluafedserxam.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62542027

A refresh, but with definitely noticeable improvements (not just a "rebadge"). RyZen 1xxx was basically beta-testing platform aimed at early adopters, while RyZen 2xxx will be fully matured product which any person could use without any worries. No segfaults, stable high clocks, even lower power envelope and temps, etc.

>> No.62542048

>clockspeeds won't be improved much
Going from 14nm to 12nm guaranties at least +400MHz stable clocks to what RyZen has right now, meaning that 100% stable 4.2GHz on all 8 cores & 16 threads will be an absolute breeze, and if you win lottery you'll be able to 100% stable 4.4GHz just fine. That is a pretty solid increase in performance in comparison to current RyZen.

>> No.62542065

On the scale from 1 to JUST, this is JUST.

>> No.62542144
File: 83 KB, 1080x1080, 1501665382820.jpg [View same] [iqdb] [saucenao] [google] [report]

fuck this is confusing
does the improved Zen 1 CPUs exist or does Zen+ == Zen2?

>> No.62542158

What's confusing about fucking Zen refresh?

>> No.62542167

so it is 100% Zen refresh?

>> No.62542170

Zen2 on 7nm comes Q1 2019, with mobile parts coming out possibly Q4 2018.

>> No.62542174

Zen+ is Zen+, a slight improvement due to shrinking of already established process, Zen 2 is Zen 2 - a new process altogether.

>> No.62542204

Son if you think this is confusing, you need to put down the weed and exercise

>> No.62542230
File: 155 KB, 804x978, Truth.jpg [View same] [iqdb] [saucenao] [google] [report]

He's just a confused Intlel shill, because how couldn't he be when this http://i.imgur.com/K92QwvT.gif is going on? And also pic-related. There's no confusion with AMD, because AMD doesn't do such shit.

>> No.62542437

>Intel Jews out with 14++: it's another money scheme
>Amd does same but calls it ((12nm)) :OMG thanks amd you sre so good

It's pretty obvious 7nm/10nm nodes will still take time to arrive. Amd is delivering 2019 at earliest on 7nm.
So called 12nm will be a mediocre update.

>> No.62542454
File: 18 KB, 488x301, images (39).jpg [View same] [iqdb] [saucenao] [google] [report]

Amd always said the zen would be 7nm.
Said it before anyone expecting 7/10 nm on a desktop chip before 2019 is drinking too much marketing coolaid

>> No.62542485

There is no such thing as Zen+, stop posting your head canon.
Zen+ was a placeholder name. Lisa Su stated in an investor call that Zen+ and Zen++ were simply called Zen2 and Zen3 now.
There is no separate Zen+, there are no other arch changes coming.

>> No.62542528

If you want to know the company that is losing, just check the temps.
Shit is always pushed up when in despair mode.

>> No.62542539


Do you think AMD could try to get an EPYC refresh on 7nm out faster though? 7nm is entering RISK production soon (1 H 2018), and AMD has leapfrogging design teams right now. Zen 2 has been worked on for a while now, and AMD might want to try sealing the deal on a market lead for a generation. If they got 7nm engineering samples say the begining of the new year, would that be enough time to validate and all that shit? To get Zen2 Epycs ready to fire by possibly Q4 2018 when 7nm might enter volume production? Maybe not though, they'd want to have mobile chips or something as the proverbial pipe cleaner first. . . Well we will see.
It might be too much for AMD to do, but I think they have a solid CPU division right now , plus new infusions of cash. If they got Zen2 out a few months earlier, that would be huge for investors and the market. Intel would be on suicide watch. CPU tech is at last getting interesting again.

>> No.62542572



>> No.62542605



>> No.62542650

>AMD waits 9 years then goes full tick tock
>meanwhile intel literally does the opposite for the last 5 years

>> No.62542910

hey fuck you. I can't be down with Intel ever after that monopoly video and because they're majority jewish
asking if I could hold of buying a 1600 for hurtz

>> No.62542947

AMD will release Zen+ on 12nm LPP on a more refinined node. OC to 4.5Ghz perhaps. It will be an interim CPU until 7nm in the first quarter of 2019.

>> No.62542993

doesn't DDR4 have internal ECC?

>> No.62543026
File: 944 KB, 1228x1502, 5kl46879udy67ojykdjhgdflyujo96uo7.png [View same] [iqdb] [saucenao] [google] [report]

Sure, Brian, sure. Shh, shhh...

>> No.62543035

Was meant for >>62542485.

>> No.62543051

>AMD does same
It's not same. 12nm is actually smaller/denser.

>> No.62543063

>AMD is delivering 2019 at earliest on 7nm
They're already finalizing 7nm at the end of this year and mass production starts at 2-nd quarter of 2018, you dumb fuck.

>> No.62543086

I bet that at this current time Intlel's temps are measured through Brian's and Paul Otellini's rectum.

>> No.62543105

7nm Zen 2 is coming at the second half of 2018, not at 2019. They're already finalizing it at the end of this year.

>> No.62543118

>He's this immature and inexperienced on the subject


>> No.62543362

Thats purely factual, you retarded little /v/irgin.
There is no such thing as Zen+. The next iteration in microarch is Zen2.

A refresh is not new arch.

>> No.62543912
File: 380 KB, 6480x2160, Zen roadmap.png [View same] [iqdb] [saucenao] [google] [report]

>There is no such thing as Zen+
Oh I'm laffin'

>> No.62543917
File: 84 KB, 253x239, 1504811041382.png [View same] [iqdb] [saucenao] [google] [report]


>> No.62543946
File: 24 KB, 267x297, 1504560195953.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62543968

I think AMD just used "Zen+" as a provisional label for future iteration(s) of Zen. I think they've stopped using Zen+ in their stuff since they've disclosed Zen3

>> No.62544046


>> No.62544216

gozaimasu keller-san desu

>> No.62544271
File: 422 KB, 672x794, 052.png [View same] [iqdb] [saucenao] [google] [report]

>> No.62544292

>"source" is a powerpoint mockup

literally fake news until confirmed otherwise

>> No.62544491
File: 971 KB, 484x682, Inturd.png [View same] [iqdb] [saucenao] [google] [report]

This is the state of Inturd's shills in 2017, lol

>> No.62544761

People use zen+ for the 14nm+(now 12nmLP) refresh, not for Zen2 or some new urach

>> No.62544920


>> No.62545409

>M.2 shield
2nd worst meme right after RGB

>> No.62545466

Only if it's MSi.
There's nothing wrong with Gigabyte's shields.
MSi's ones were trash, but that was long ago and no one does same mistakes now.

>> No.62545482






>> No.62545493

>being this much of a retard
Exactly as I stated, Zen+ was a placeholder name. Su confirmed in an investor call that the successive arches in the Zen family are Zen2 and Zen3.

There is nothing called Zen+. Stop being a shitskin retard.

>> No.62545589

Sheesh, fine fine, I admit your defeat if you're wishing for it so much.

>> No.62545685
File: 74 KB, 700x362, roadmap.jpg [View same] [iqdb] [saucenao] [google] [report]

stupid bonehead, kill yourself

>> No.62545971

Cannonlake was supposed to be out in 2016, the 10nm node just sucks because apparently Intel has fired all their competent engineers.

>> No.62546012

Don't forget that there is a version that supports ECC (Xeon W-2195), but that's $2550 and requires the even more expensive enterprise chipset to even run.

Also interestingly enough, the Xeon-W line has a 4c/8t Skylake-X chip that does have the full 48 PCI-E lanes and quad channel MC, it's only the consumer lineup that has to make do with a repackaged Baby Lake for the entry-level chips.

>> No.62546042
File: 971 KB, 484x682, JewFail.png [View same] [iqdb] [saucenao] [google] [report]

>10nm node just sucks

>> No.62546051

Sorry, was the sentence too long for you to comprehend?

>> No.62546071
File: 1.13 MB, 232x181, Curry, but not Pajeet.gif [View same] [iqdb] [saucenao] [google] [report]

Just one jew killer >>>>> 100000 "Intlel's (((((competent))))) engineers"

>> No.62546086
File: 112 KB, 691x771, 1499963016439.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62546116

>I'm too stupid to understand the word "because"
You almost make me WISH I was defending Intel.

>> No.62546135


>> No.62546177
File: 1.25 MB, 1002x1480, t.JUSTnich.png [View same] [iqdb] [saucenao] [google] [report]

>Just one jew killer >>>>> 100000 "Intlel's (((((competent))))) engineers"
Just one Contract Keller vs. 12Bn $ R&D
how can a Company fuck up so damn much

>> No.62546190
File: 11 KB, 239x211, dea216e1ee39e437d104dab89311f611.jpg [View same] [iqdb] [saucenao] [google] [report]

Meanwhile I overclocked my Ryzen 5 1600 to 4 Ghz with the stock cooler and it works as well as the i7 7700k for 200$ with a lower temperature as well.

>> No.62546282
File: 3.33 MB, 470x264, OH YISS.gif [View same] [iqdb] [saucenao] [google] [report]

And RyZen+ will easily do 4.2GHz on air, with going up to 4.5GHz if lottery is won. With all 8 cores and 16 threads enabled. While being even cooler and consuming even less power than RyZen does right now. First quarter of 2018, babeh. And same socket, so changing RyZen 1xxx to RyZen+ 2xxx will be easy as fuck. OH YEAH, MOTHERFUCKERS, OH YEAH!

>> No.62546346
File: 743 KB, 1126x1091, 714.jpg [View same] [iqdb] [saucenao] [google] [report]

>that guy who bought an overpriced LGA1151 quad core processor on the chopping board
>in 2017

>> No.62546357
File: 190 KB, 500x733, Turned.png [View same] [iqdb] [saucenao] [google] [report]

>Just niche
The state of Infail circa 2020~, lel.

>> No.62546392

Wrong face.
Needs Otellini/Krzanich/Ryan Shrout/JayzTwoCents/GaymerNesux/Linus/LPT's Lavco.

>> No.62546467

How many billions do you think intel is going to waste on doing illegal shit again?

>> No.62546481
File: 64 KB, 600x800, AD.jpg [View same] [iqdb] [saucenao] [google] [report]

>same socket, so changing from A to B will be easy as fuck
This is the one thing Intel ABSOLUTELY cannot allow in this modern day and age. Last time this was with Z77 and Sandy-to-Ivy.

>> No.62546543

Back then internet wasn't as developed and not many people used it in general, so that shit worked. Nowadays it's almost impossible, so only thing Inturd is capable of doing, is such dirty "tactics" as this http://www.techpowerup.com/235092/intel-says-amd-epyc-processors-glued-together-in-official-slide-deck . And it actually backfired big-time on them, so they've gone silent as of right now, weeping away in a corner like the prison rooster bitches they are. JUST-ICE.

>> No.62546554

But muh sysmark

>> No.62546563


>> No.62547082
File: 137 KB, 356x438, qodf2dy.png [View same] [iqdb] [saucenao] [google] [report]


>> No.62547569
File: 28 KB, 394x367, oh my2 .jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62547799

better enable those fucking prim shaders fucking pajeet codemonkeys

>> No.62548168
File: 20 KB, 320x306, 53hnhlfdtjkltdfjgldgkdot6098546u45uj756rj.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62549887


>> No.62550153
File: 90 KB, 769x1024, BYsxpQJCUAAo6ix.jpg large.jpg [View same] [iqdb] [saucenao] [google] [report]

How can other companies even compete?

>> No.62550193
File: 84 KB, 640x429, 1496686124814.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.62550218
File: 45 KB, 778x512, 1498104532557.png [View same] [iqdb] [saucenao] [google] [report]


>> No.62550360
File: 1.17 MB, 940x433, Legitimate Intel Fire Group.png [View same] [iqdb] [saucenao] [google] [report]

There's something burning in your background, here, let me help you with that

>> No.62550939

Why don't they glue their cpu and BTFO amd?

>> No.62550972

because they can't

>> No.62551409

Have they bothered to actually put effort into their bioses?

>> No.62551481

>perfect on all fronts
>8+4 cpu power
no thanks

>> No.62551551

>core values: credibility
Somehow I think this might not be the Intel we are talking about ITT

>> No.62551587
File: 207 KB, 480x468, 1496606603102.png [View same] [iqdb] [saucenao] [google] [report]

> mfw one's kikery blowns themselves out.
I wonder how many engineers and resources they waste coming up w/ these artificial segmentations for shekels.
> mfw your shekel seeking schemes result in soo many segmented architectures that you can't even reorganize yourself and present a cohesive defense against your competitors once they've blown you the fuck out.

>> No.62551618

Can anyone tell me why intel still heavily invests in its own fabs/fab tech when its clear that other foundaries do it cheaper, better, faster? They're fumbling around w/ 10nm and the industry is on 7nm... wtf

>> No.62551754

no, the zenith is still easily the best

>> No.62551929

their shares will plummet if intel says that their biggest game isn't paying off and they're shutting it down to use 3rd parties

>> No.62553506
File: 1.28 MB, 320x240, 1439337735449.gif [View same] [iqdb] [saucenao] [google] [report]

There are signs that the "14nm+" products for 2018 got renamed to 12nm, because GloFo decided lower number for polished 14LPP process sounds better.

>> No.62553511

mid-2019, more realistically.

>> No.62553550

It is not clear they are cheaper. Intel probably enjoys massie economies of scale and gets to keep much more money than AMD from a comparable chip.

Look at it this way - from AMD's sale, the margin has to feed both AMD and GlobalFoundries. Intel pockets both shares for itself.

>> No.62553632

it actually has some area reduction, so it's not completely unwarranted

>> No.62553643

Intel also throws billions worth of R&D money at fabs every fucking year. Over 50% of their total R&D budget is fabs.

>> No.62553659

The area reduction comes from denser (7.5 tracks) SRAM.

>> No.62554239

>How many billions do you think intel is going to waste on doing illegal shit again?
I honestly don't think Intel can pull that off again.
1. The flow of information is insane these days, bad PR would harm Intel.
2. Most companies are sick of Intel at this point, and jumped ship as soon as TR and eypic were announced.
I think it was HP who couldn't even function without Intel's bribes.
They definitely don't want to go down that road again.

Name (leave empty)
Comment (leave empty)
Password [?]Password used for file deletion.