Quantcast
[ 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: 402 KB, 743x480, IntelonSuicideWatch.png [View same] [iqdb] [saucenao] [google] [report]
58374284 No.58374284 [Reply] [Original] [archived.moe] [rbt]

IT'S OVER INTEL IS FINISHED!

>> No.58374348

>>58374284
This thread again. Jannies, for once, do your fucking job.

>> No.58374352
File: 194 KB, 500x500, stu.jpg [View same] [iqdb] [saucenao] [google] [report]
58374352

>his CPU isn't soldered to the board and held together by gorilla glue
enjoy your placebo AMDfags

>> No.58374376

>>58374348
It's a new article

http://wccftech.com/amd-confirms-ryzen-cpus-unlocked-overclocking/

First we have a four-year architecture, and now this.

>but that reminds me of Bulldozer

based Keller and Lisa Su is at the right time brother.

>> No.58374392

>>58374376
>poointhelootech
No. Someone should make a proper Zen thread with proper OP instead of shitposting magnet.

>> No.58374410
File: 210 KB, 1280x706, 1480895518910.jpg [View same] [iqdb] [saucenao] [google] [report]
58374410

Arent all amd cpus unlocked?

>> No.58374427

>>58374284
>needing overclock to defeat Intel

>> No.58374435

>>58374284
Except, you know, every single Intel CPU that it competes with is unlocked. So if you knew that, you'd be surprised if any Ryzen CPUs were locked at all.
At least before the mainstream APUs are released.

>> No.58374464

>>58374410
And have been since forever, and was one of the things that drew attention to them way back in the day when they weren't irrelevant.

>> No.58374557

>>58374427
>Preferring a locked CPU to an unlocked CPU

>> No.58374590

>>58374557
You can get a cheaper motherboard if you don't want to OC.

>> No.58374611

when AMD has no budget and can't afford to install extra lockout circuitry

wew

>> No.58374625
File: 418 KB, 627x627, 1481671240637.png [View same] [iqdb] [saucenao] [google] [report]
58374625

>>58374284

>> No.58374628

>>58374284
Both Intel and Nvidia will remain the top dogs in their domains for a long time.

>> No.58374637

>>58374625
I've been waiting for this picture again

>> No.58374647
File: 527 KB, 543x810, 1343390600839.jpg [View same] [iqdb] [saucenao] [google] [report]
58374647

B1 stepping(F4 OPN) clocked 3.6ghz base, 4ghz peak single core turbo.
Overclocking Ryzen is controlled by the mobo, all CPUs by default are unlocked, they just ensure junk tier budget boards don't support overclocking so theres no VRM throttling and power delivery issues.

Everything so far is looking really good for Ryzen. Competition in the desktop CPU space, and 2017 as a whole is already off to a fantastic start in the rest of the market. RAM, storage, GPUs, displays, every segment is looking very promising right now.
Its shaping up to be the best year for hardware that I can remember in my 30 years of life. The future is looking bright and comfy.

>> No.58374648
File: 39 KB, 499x449, 077310.jpg [View same] [iqdb] [saucenao] [google] [report]
58374648

Hahaha this thread is reminding me of Bulldozer launch.

Good one OP

>> No.58374654

>>58374464
Only Intel "K" model chips are.

Pretty much all AMD chips have been for the past 5-10 years.

>> No.58374658

>>58374625
Keller left AMD nearly a year and a half ago

>> No.58374663

>Ryzen will launch closer to the middle of Q1 or early Q1 rather than the end of Q1. AMD wouldn’t give us a specific date but sometime in February is a good bet.

>> No.58374685

>>58374658
He did his job. He's batman of CPU world.

>> No.58374717
File: 61 KB, 600x600, 1441830753495.jpg [View same] [iqdb] [saucenao] [google] [report]
58374717

>>58374654
This isn't true, some mid-high end AMD APUs and AM3+ low end has a locked multi.

>> No.58374770

>>58374663
The same auther on wccftech said late 2016 was a good bet for Zen as well
why would you trust speculation, is there *anything* to support their claim of middle Q1 for ryzen ?

>> No.58374837

>>58374770
I know anon, I'm just fooling myself desperately grasping one rumor to hold onto.

I know curryttech isn't thrustworthy but I'm finding myself going back there almost daily since I boarded the Zen hype train.

>> No.58374838

So X370 is more limited than Z270 and that's supposed to be AMD's enthusiast platform while Z270 is an Intel mainstream platform?

>> No.58374866
File: 108 KB, 1024x724, 4iivjn.jpg [View same] [iqdb] [saucenao] [google] [report]
58374866

>>58374770
Morons, including those of wccftech, kept saying October 2016 for Summit Ridge because of this slide in conjunction with statements made by Su and Papermaster.

"IP" on this slide means final silicon is in production, which it is. AMD hit all their milestone here. A0 steppings were in production as early as June, the latest steppings were produced in October and November 2016.
Su and Papermaster both mentioned limited availability end of 216, with full availability Q1 2017. They stuck to that as well, limited availability has been OEM sampling. Aside from the leaked engineering samples that everyone has been seeing, AMD sent out OEM samples to tons of vendors clocked at 3.2ghz/3.5ghz.

Apart from this, theres been no talk of dates. Just Q1 2017. Beyond that everyone is guessing.
We do however know that there will not be a paper launch. They recently promised full availability as soon as its announced.

>> No.58374903

>>58374376
>wccftech
fucking street shitter. stop spamming your faggy site. it's just a mill for shit to bait retards without adblock.

>> No.58374904

>>58374838
Zx70 are Intel's enthusiast platform.

>> No.58374920

>>58374904
X99 is Intel's enthusiast platform.

>> No.58374936

>>58374920
Both are the only ones that support OCing.

>> No.58374965

>>58374838
Kaby and Z270 doesn't even have native USB 3.1 Gen2.

>> No.58374987

>>58374648
nah, there were several hints and outright misdirections (where's Jeff I wonder) that made any tech forum reader suspect that there could be something wrong with bulldozer.

The hype is at the same level if not greater this time, but different than with Buldozer AMD learned how to control the leaks very effectively, especially those that could give rise to bad press or detrimental wild rumors.

>> No.58375016

>>58374410
There are locked AMD CPUs but the price is so low anyway it makes no sense to buy the locked one instead of the already cheap unlocked ones.

>> No.58375062

>>58374376
>wccftech
Filtered.

>> No.58375070
File: 37 KB, 175x132, jesus1.png [View same] [iqdb] [saucenao] [google] [report]
58375070

>>58374647

>mfw we've hit the limit in terms of CPU power

>> No.58375087

>>58374838
these pruducts haven't even launched ffs.
Anon only saw a couple of teasers for clearly yet unfinished motherboards.
Zen is a flop.
kys.

AM3 had a single "enthusiast" class motherboard with the Asus Crosshair for 2 or 3 years when msi and gigabyte decided that even that tiny slice of market wasn't something they weren't comfortable with allowing an Asus effective monopoly. The fact that we're seeing premium SKU's that are comparable with intel chipset, even in the branding names, ready to launch with Zen from the onset is actually giving credence to the fact that AIB's are onboard the hype train.

>> No.58375123

>>58374284
MUH GIGAHURTZ

Yeah, that worked so well for bulldozer and vishera

>> No.58375140

>>58375087
It's not even out yet. How's it a flop?

>> No.58375157
File: 543 KB, 1100x1002, 1471589730534.jpg [View same] [iqdb] [saucenao] [google] [report]
58375157

>>58374376

>> No.58375182

>>58374987
That cunt JF-AMD posting on overclock.net was doing legitimate damage control. He was directly contradicting official statements made by AMD during their 2010 HotChips presentation. The icing on the cake was when he said something like "people don't care about power consumption, they just want to plug it in and get their performance." People could tell there was something wrong.
The desperate upset about Zambezi's launch was fanboys thinking some magical BIOS or AGESA update would unlock more performance and change bench results.

The lead up to Ryzen's launch is nothing like what we saw with Bulldozer. The Zen HotChips presentation was overwhelmingly positive, AMD is talking about getting back into high performance. In 2010 AMD was making excuses, admitting there was a major downside to the arch, but they were working on it.

All the diehard intel fanboys are just running out of straws to grasp here.
At first they tried to claim that the 40% figure was a lie.
They tried to claim that a 40% uplift wouldn't even reach Sandy Bridge performance.
They tried to claim that 14nm LPP was a terrible process.
They tried to claim that Zen would never clock over 3ghz, and have no OC headroom.
They tried to claim Summit Ridge would only support single channel memory.
All of that shitposting FUD has been shotdown, and now they basically have nothing left, so they're all running away and hiding on the forums.

When the CanardPC review came out and they explicitly said that Zen was near Broadwell IPC a few people tried to make the clock speed argument, but that fell flat instantly. Su guaranteed 3.4ghz+ at launch, and they had 3.6ghz/3.9ghz CPUs running at CES, and now we know that there are 3.6ghz/4ghz chips finalized.

Ryzen looks good all around. Its nothing like Zambezi.

>> No.58375197

>>58375123
Actually it would have if the fabs delivered. As much of a mess as Bulldozer was it was intended to run @ 5+ ghz but gofo never delivered on 20nm.

>> No.58375223

>>58375182
not overclock, xtreme systems.
t. real oldfag.

OC.net was the kiddies table back then.

>> No.58375236
File: 278 KB, 500x382, 1358805432312.png [View same] [iqdb] [saucenao] [google] [report]
58375236

>>58374987
Its the AMD boi "Intel is finished" meme thats reminding me of bulldozer.

Ryzen could be amazing, and it should be because competition is good for the consumer.

Nothing is going to happen to Intel though, and i cant really see ryzen beating high end X99 gear.

>> No.58375265

>>58375197
Lowering IPC to get 5Ghz+ was not a good idea to begin with.

>> No.58375285
File: 48 KB, 800x800, gk9MvzJ.gif [View same] [iqdb] [saucenao] [google] [report]
58375285

>>58375182
>They tried to claim that 14nm LPP was a terrible process

It is. Its vastly inferior to Intel's 14nm. Zen is just a significantly better architecture.

>> No.58375294

>>58375182
Only my gut instinct and you may call it fanboyism, but I can't shake off the feeling that AMD is still holding their biggest cards tightly regarding Zen.

>> No.58375315

>>58375294
APUs with onboard HBM memory would assravage Intel bretty hard.

>> No.58375321

When are the actual reviews going to be released? I thought it was supposed to happen during CES, but nothing still.

>> No.58375328

>>58375285
How bad is it? Anything to worry about?

>> No.58375331

>>58375197
GloFo didn't have a high power 20nm node.
The Bulldozer arch was designed around the 32nm PD-SOI process that they licensed from IBM which was tailored for IBM's POWER7+ line. AMD wanted to target clocks of about 4.5ghz. The original FX 8150 however only launched at 3.6ghz, peak 4.2ghz turbo, though it was doing so with horrendously out of spec leakage.

The FX 8350 pushed clocks a bit higher, hitting 4ghz base while managing to bring down power a bit simultaneously. They still didn't reach original goals inside of their power envelope. Some latter 8370e chips can hit 4ghz at just 90w~ though.

>>58375223
Apparently he made the same statements all over the place. If you google it he has some damage control apology on Anandtech from 2011 after the release.

>>58375285
If 14nm LPP was a bad process, it isn't, they wouldn't be reaching these clocks inside of the conservative power envelope of 95w.
"Low power" does not mean low clocking anyway. Intel's processes are "low power" in the same regard.

>> No.58375350

>>58375265
There is nothing wrong with speed demon design if you meet your clock goals.

>> No.58375445

>>58375350
Problem is future processor roadmap. You get 5Ghz and then what next, praying for better future process and 6Ghz or tweaking arch for better IPC? Getting solid IPC first and increasing clocks gradually with maturing nodes is a better overall strategy

>> No.58375477

>>58375331
Yeah, that was one freaking PR nightmare. AMD actually managed to put the lid on it quite effectively -- after the fact --, since they limited the amount of damage it could have given to their image (I'm not saying that they came out unscathed, quite the contrary, but that shit could've literally made them finished and bankrupt).

I just wanted to point out that Xtreme Systems was his main platform and where the flames reached the highest.

>> No.58375502

>>58374717

>Whats a BCLK

>> No.58375506

>>58375350
t. intel netburst engineer from the good old days.

>> No.58375507

does anyone remember what was going on pre bulldozer?
basic chronology? how does it compare?

>> No.58375512

>>58374611
When you pay $60 extra for a K version chip just to fund the R&D that cucked you into needing to buy it in the first place.

Wew

>> No.58375519
File: 28 KB, 550x550, single socket holocaust.jpg [View same] [iqdb] [saucenao] [google] [report]
58375519

>>58375506
LONGER. MAKE LONGER PIPELINE.

>> No.58375566
File: 39 KB, 432x288, butt.jpg [View same] [iqdb] [saucenao] [google] [report]
58375566

>>58375502
You can only go so far with bclk. Also iirc AMD IMC is pretty shitty so you are gonna have some ram issues if you use BCLK, but things may have changed i havent had an AMD CPU since Phenom II

>> No.58375587
File: 973 KB, 801x1500, 1430419950754[1].png [View same] [iqdb] [saucenao] [google] [report]
58375587

>>58375519

>> No.58375629

>>58375477
their stock is actually costs more now than in 2010
it's almost bankrupt them they dropped to two same year

>> No.58375644

>>58375587
Their only problem that the pipeline wasn't long enough

>> No.58375652

>>58375507
hype is greater and more mainstream, probably due to youtube techies reaching more normies.
no iffy stuff
no flaming retard running the PR / social media management

they've been feeding the press what they want and they want, so I'd say that have good control over the entailed discussions.

I think that this will be an awesome product, simply because I can't fathom that AMD got this good at PR after being veritable bottom of the barrel in anything media related since ever.

My take is that, since we still haven't heard anything bad about it, then, there probably isn't.

>> No.58375682

>>58375652
Basically, it. There's nothing bad to say about Zen unless you want free (you)s.

>> No.58375688

>>58375519
>that fucking filename
This will never stop being funny. Goddammit.

>> No.58375690

>you'll never get to unlock your CPU with a graphite pencil again

>> No.58375712

>>58375512
paying extra for a "k" is nothing. Intel actually charges you to disable architectural features on their chips. And, naturally, they double dip and charge again on the sku's that have them enabled.

>> No.58375738

>>58375712
this guy knows what's up

>> No.58375741

>>58375688
Yes. Because it was fucking truth. NetBurst was the worst thing Intel has done in forever.

>> No.58375748

>>58375315

They are going to be scary.

https://youtu.be/KdghvnwIvv8

Forza on integrated graphics, 1080p mix of ultra and high at 30FPS. And it's a shitty excavator with DDR4.

Zen mobile APUs with HBM, will fuck iris pros.

>> No.58375751

>>58375587
>>58375519
explain

>> No.58375758

>>58375506
I can't now find it where but one Intel engineer said he doesn't regret P4. They had to find ways for better power management and increasing efficiency with good hyperthreding, they later used those for Core architecture for great success. Same goes for AMD, Keller said something similar. They had the basic building blocks and technology for Zen already, they had developed good power management and turbo for Excavator and fixed the bottlenecks of bulldozer in Zen

>> No.58375768

>>58375712
What architectural features?

>> No.58375784
File: 944 KB, 800x600, 1379307297115.png [View same] [iqdb] [saucenao] [google] [report]
58375784

>>58375690
Good shit.

Better times.

>> No.58375790

>>58375587
To this day I think that the AMD house fires meme is an concerted effort from salty old Intel fanboys from the netburst days and Nvidia's fermi.

>> No.58375815

>3.6/4.0Ggz 8 core ES within 95W
>still not final revision
>500$

intel status = total anal devastation

>> No.58375817

>>58375751
hownew.ru
>>58375768
HT i think.

>> No.58375829

>>58375751
https://www.youtube.com/watch?v=7uBNCN6v_gk

>> No.58375878

>>58375748
>Zen mobile APUs with HBM, will fuck iris pros.
Shit, iGPU that would be able to do gaymen.

>> No.58375891
File: 498 KB, 320x240, 1393382507546.gif [View same] [iqdb] [saucenao] [google] [report]
58375891

>>58375790
Idk man the 290x ran just as hot as fermi

>> No.58375904

>>58375790
original housefire meme was about intel and nvidia

propaganda machine reversed it, no doubt

>> No.58375905

>>58375878
>>58375748
well if they're paying enough attention to license onboard graphics from AMD, maybe

>> No.58375908

>>58375817
several hardware level encoders that trickle down from the xeons, architectural integrity, that are too good for the plebeian tier i# cores.

>> No.58375911

>>58375790
Last gen AMD cards (anything that isn't Fury or RX 480) unironically consume more power than GTX 480. Never mind GTX 580, which is the refined Fermi.

So if AMD being housefire is a meme, so is the GTX 480 being a housefire. Oh and the FX 9590 consumes more power than any CPU intel has ever produced, including overclocked Pentium EEs. So yeah, if you want a housefire, AMD is your go-to.

The only saving grace AMD has is that coolers today are better than they were when Fermi was released and when Pentium D was the high end. So AMD housefires have more copper and aluminium dissipating their massive heat output.

>> No.58375918

>>58375891
on stock cooler, difference is fermi ran hot on partner coolers

>> No.58375932

>>58375815
It is really exciting to see.
I wonder if they'll release another stepping later into the year, or have an even higher clocked refresh in 2018. With base clocks this high this early its definitely within the realm of possibility.
These things are going to underclock/undervolt like gods. They have to be hitting 3ghz~ at just 65w.

>> No.58375933

>>58375911
consequently, instead of a housefire, the R9 fury is an efficient HVAC solution

>> No.58375936

>>58375790
So are you implying that AMD has no housefires right now? When literally every single CPU and GPU from their line runs hotter and consumes more power while at the same time delivering less performance than any Intel/Nvidia CPU/GPU equivalent?

Also, do you know about a CPU called FX-9590? Did you hear about it right? I'm pretty sure you know it needs a FUCKING WATERCOOLER to not burn your whole house down?

But no, you won't admit anything of this, you would just call me a Nvidia/Intel jew shill xD.
Right? xD !!!1!

>> No.58375948

>>58375936
You sure are buttdevastated.

>> No.58375984

>>58375911
Yes anon, copper and aluminum suddenly got better in 2014.
Nvidia and Intel marketing redfined the very laws of physics and did a damn good job outdoing God himself in material engineering.

See, the salt lasts for decades.

>> No.58375995

>>58375948
Insightful retort

>> No.58376003

>>58375948

Thanks for exactly proving my point.
Have a good day Pajeet!

>> No.58376016

>>58376003
A нe пpoшёл бы ты нaхyй, пeтyшoк.

>> No.58376019

>>58375905

Intel is all in on nvidia licenses and using edram. They would have to start completely from scratch to integrate AMD technology in their iGPU.

>> No.58376032

>>58375815
TDP doesn't matter.

>> No.58376037

>>58376016

Sorry Ivan, I don't speak moonrunes.

>> No.58376041

>>58376019
they probably have a plan if they moved to AMD for patent deal

>> No.58376048

>>58376037
Then fuck off. /g/ is rightful russian clay.

>> No.58376065

>>58376032
overclocking an 8 core Intel leads to 200W power consumption and unreasonable temperatures.
Overclocking Zen will go much easier

>> No.58376073

>>58375918
http://www.guru3d.com/articles_pages/msi_geforce_gtx_480_lightning_review,8.html

Oooh, so hot. Know what else runs at 74 degrees? A MSI GTX 980 4G Gaming which has a TDP of around 200W. Which as far as high end GFX cards go is quite low.

>>58375984
You fucking retard. Coolers were cheaper and shittier in 2009/2010 than they are today. There were good coolers like MSI Twin Frozr, but it was less commonplace than it is today and that cooler is nowhere near as good as the Twin Frozr V used today. Even reference coolers are way better today, never mind the fact that in 2009 things like cards turbo boosting or throttling back clocks to keep within thermal envelope was unheard of. The GTX 480 ran full blast even in furmark even though modern cards like R9 290X or GTX 1080 throttle to shit in it.

>> No.58376090

>>58376065
>overclocking an 8 core Intel
Is impossible unless it's a sandy bridge.

Everything from Ivy to Skylake overclocks like crap, only Kabylake got back to Sandy levels of overclocking and Kabylake-X won't be here until middle of 2018

>> No.58376105
File: 63 KB, 421x248, 1726450i03g9jmahjj0mjj.png [View same] [iqdb] [saucenao] [google] [report]
58376105

>>58375936
>delivering less performance than any Intel/Nvidia CPU/GPU equivalent?

Why would someone do this? Just go on the internet and tell lies?

The 7970 was VASTLY superior to the GTX 680 and the 290x stomps on the 780ti. AMD's performance was so much better slightly tweaked versions of the 290/x (390/x) now beat cards of a gen newer design. Granted they do so by channeling the raw power of the sun's core but they do it.

>> No.58376115

A NEW SOCKET A YEAR KEEPS THE AYYMD GOYIMS IN FEAR

>> No.58376124

>>58375984
>no cubic boron arsenide heatsink
>can't keep 200w chip near room temp with a block small enough to fit in the palm of your hand

Material research needs to reach market faster.

>>58376019
Intel's IGP arch isn't based on any Nvidia IP. What they licensed is a patent that basically describes what a GPU fundamentally is in the most abstract way possible. Their licensing deal just ended, and intel just struck the same deal with AMD to the tune of several billion dollar over a few years.

>> No.58376139

>>58374376
>>58374376
>currytech

>> No.58376180

>>58376048
/г/ цe нe pyccкi, a пyтiн хyйлo! кypвa!

>> No.58376200

>>58376124
>Material research needs to reach market faster.
there is better thing in the works - themoelectric freezers
basically semiconductor cooler, no moving parts, no compressor, no coolant liquid

>> No.58376227

>>58376200
Sounds like another meme waiting to happen.

>> No.58376233

>>58376180
Цe нe yкpaїнcькa. Цe якecь лaйнo. Bбийcя, бyдь лacкa.

>> No.58376239

>>58376227
i mean, peltier coolers were tried a few years back, and they had a few.....problems

>> No.58376240

>>58376065
>overclocking an 8 core Intel leads to 200W power consumption and unreasonable temperatures

Nigga what i have a 6900k OCed at 4.5 can get by on a Hyper 212.

On water it clocks at 5.1 and my loads are 45c

>>58376090
Keep telling yourself that boyo.

>>58376105
>and the 290x stomps on the 780ti

They are verrry close within the margin of error. No stomping ether side.

Damn the AMD blind fanboys are out in force.

>> No.58376258
File: 9 KB, 247x281, 1397431335808.jpg [View same] [iqdb] [saucenao] [google] [report]
58376258

>>58376240
http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/70125-gtx-780-ti-vs-r9-290x-rematch-3.html

Extremely close.

>> No.58376265

>>58375790
Many people were calling AMD running hot in the Athlon XP days, it didn't had a heatspreader and it was hard to properly fit a HSF on exposed die. There was no auto TDP management those days so some improperly seated chips literally burst into flames or had cooling problems.
https://www.youtube.com/watch?v=NxNUK3U73SI

AMD experience was not good in K6/XP days because of the non performance reasons, it was hard to find good motherboards and chipsets were shit. Ironically now Nvidia had the best support for AMD, their drivers supported 3DNow! acceleration and nForce was the best chipset for the platform.

>> No.58376266

>>58376240
>Keep telling yourself that boyo.
I keep telling myself how everyone and their mother complained to Intel about shitty overclocking post SB?

Yeah, I'm delusional.

>> No.58376284

>>58376124

I forgot they went full patent troll.

>> No.58376299
File: 959 KB, 264x248, 1395807722866.gif [View same] [iqdb] [saucenao] [google] [report]
58376299

>>58376266
Overclocking is pure luck. Sure stuff past SB dosnt clock as high as OFTEN

>> No.58376321

>>58376258

I trust linus over Hardware Canucks. That eurotrash wannabe with soft focus filters doesn't know a fucking thing

>> No.58376334

>>58376124
>ntel's IGP arch isn't based on any Nvidia IP. What they licensed is a patent that basically describes what a GPU fundamentally is
they paid 600m yearly for it though

>> No.58376348
File: 59 KB, 337x367, 523565.jpg [View same] [iqdb] [saucenao] [google] [report]
58376348

>>58376321
>i trust linus.

That motherfucker makes clickbait and you trust him?

Also other sources have the same benchmark results between the two, they are extremely close. Neither is faster overall.

>> No.58376367

>>58376334
A fee for avoiding lawsuits is nothing to a $50 billion per year company.

>> No.58376401

>>58376227
>>58376239
corporations seem to invest in it for real this time
some material workaround for better efficiency or something happened

>> No.58376415

I just bought an i5 6600k and a z170 board cause my fx8350 @4.7 Gigalulz was bottlenecking my rx480 in some games, i really think i made a good decision, if AMD fucks up again Intel might be increasing prices.Better play safe.

>> No.58376440 [DELETED] 

>>58376321
>Linus
>A known Nvidia shill

https://www.youtube.com/watch?v=CMTUZtF27kA

Amazing job, anon.

>> No.58376459
File: 1.66 MB, 750x1334, IMG_0107.png [View same] [iqdb] [saucenao] [google] [report]
58376459

>>58374284
Does this mean I should hold off from purchasing a laptop?

>> No.58376465

>>58376459
Yes. AMD might present APUs with onboard HBM memory.

>> No.58376473
File: 44 KB, 500x564, 1384747069383.jpg [View same] [iqdb] [saucenao] [google] [report]
58376473

>>58376415
Something tells me that combo wouldnt cause a bottleneck.

6600k is a great cpu though and should be a good for a long while. depending on what you are doing.

>>58376440
No use trying to reson with cult fanboys. of "my company can do no wrong. i must defend my purchase"

>> No.58376494

>>58376265
christ why is that video so fucking blocky? Looks like it's being played with VLC

>> No.58376503
File: 338 KB, 500x376, 1380102773110.png [View same] [iqdb] [saucenao] [google] [report]
58376503

>>58376465
Anyone who says you should hold off for "X" is talking out there ass, tech always moves forward just make your purchase.

>> No.58376510

>>58376503
>tech always moves forward just make your purchase.
Just like Kaby Lake.

>> No.58376522

>>58376258
http://www.babeltechreviews.com/nvidia-forgotten-kepler-gtx-780-ti-vs-290x-revisited/3/

>> No.58376538

>>58376494
Video is from 2001, people were still on dial-up those days so they had to keep bitrates small. It was later reuploaded to youtube

>> No.58376559

>>58376510
>8k decoding
>better iGPU
>more efficient

Only people mad at Kaby Lake are the 'muh performance' cucks

>> No.58376560

>>58376503
Dumbass.

>> No.58376570

>>58376559
>a better iGPU
No one with a discrete card cares, fucking no one.

>> No.58376581

>>58376503
that's not why people are saying to wait here, it's about ryzen possibly affecting pricing, not so much raw performance

>> No.58376587

>>58374284
You left out that you have to by a motherboard amd's high end chipset to overclock the cpu

>> No.58376598

>>58376587
No? B350 supports OCing.

>> No.58376607

>>58376570
laptops with discrete GPUs are for children.

>> No.58376623

>>58376587
[citation needed]

>> No.58376627
File: 61 KB, 625x552, 1349851775418.jpg [View same] [iqdb] [saucenao] [google] [report]
58376627

>>58376510
What i said has nothing to do with companies fanboyism or anything.

If you wait for "x" why not wait for "y" and then wait for "z"

Honestly with the amount of advancement made between generations, waiting isnt really worth it.

>>58376522
>from an tech site ive never heard of.
>still pretty close, with slightly inflated numbers from other review sites

Really makes you think.

>>58376560
/g/ has the best insults, super creative.

Seriously, why not just wait for the holy grail then.

>> No.58376636

>>58376607
>laptops

>> No.58376703

>>58376627
>Never heard of

That's too bad, anon.

>Still pretty close
>The 290X beats the 780 Ti in its own games

>> No.58376710

>>58376473
I was tired of waiting, waiting for a release day for Ryzen, then waiting some more for real world benchmarks gaming benchmarks from non biased tech reviewers, comparisons and such, i make both gaming and some heavy browsing, mostly that, host some LAN games but my fx was going very low on fps on certain games like just cause 3 and gta v, i think i am good to go for a good time.

>> No.58376751
File: 185 KB, 420x343, 1349394174256.png [View same] [iqdb] [saucenao] [google] [report]
58376751

>>58376703
>OMGZ it beats it SOMETIMES by sometimes less than an entire frame.
> No frametimes

>Yeah too bad ive never heard of your rinkydink tech review site.

Yeah suck that corporate cock. AMD, Intel and Nvidia love the free advertising.

>> No.58376826

>>58376581
>>58376503
not only that, but when you are right at the launch of a new tech thing, building an entire computer over something like intels kaby would be stupid

unless you can not wait, your computers dead, you are working professionally, you have a 3 month window for a potential game changer.

>> No.58376847

>>58376623
to be fair, you need good power delivery to oc a cpu by any meaningful metric.

>> No.58376889

>>58376751
Why are you so heated up, anon? This is a civil discussion, isn't it?

If you provide me a website with that many games, then i'd be happy to take a look at it.

>15FPS difference in Star Wars: Battlefront
>20FPS difference in Batman: Arkham Knight, an Nvidia GameWorks title
>Also beats it in multiple other GameWorks titles

C'mon, anon.

>> No.58376918

>>58376847
Only if you buy AMD, Intel doesn't have this shit.

>> No.58376945
File: 46 KB, 604x452, Capture.jpg [View same] [iqdb] [saucenao] [google] [report]
58376945

I've been needing to upgrade for some time now, what should I wait for? Zen? Vega?

Even if I go Intel/Invidia, should I wait for those to come out in hopes of prices dropping?

>> No.58376959

>>58376945
also why do my two HGST's show up with slightly different names?

>> No.58376974

>>58376945
You can go for Zen/Vega, Zen/Pascal, Zen/Poolaris, whatever, dude. But you'd certainly wait for Zen.
>>58376959
Speccy is dumb.

>> No.58376997
File: 45 KB, 437x471, excited loli.jpg [View same] [iqdb] [saucenao] [google] [report]
58376997

>>58374685
> He's batman of CPU world.

Does that mean he's working on the Batmobile in some underground cave for Tesla?

>> No.58377026
File: 67 KB, 520x853, 1349726471431.jpg [View same] [iqdb] [saucenao] [google] [report]
58377026

>>58376826
I would understand if it was a new D3D or render API and you need HW to support it. But that hasn't happened since the launch of DX11 The days of breakthrough performance in one generation difference are gone.

>>58376889
Im heated because all /g/ has become is HW shit flinging.
They are extremely close, with some titles going one way or the other. Just like its been for over 10 years.


Why even post that if its a civil discussion? You are just going to start shitflinging

>asks why someone is heated
>entire post was bait

>> No.58377051

>>58377026
>Implying /g/ was any good to begin with

>> No.58377067

>>58376945
Better just wait™ if you'd made it this far

>> No.58377080
File: 173 KB, 519x384, 1356506520357.jpg [View same] [iqdb] [saucenao] [google] [report]
58377080

>>58377051
It wasn't always this bad, but that could be said for 4chan as a whole.

>> No.58377092

>>58376974
>>58377067
Will Zen cause Intel prices to fall?

>> No.58377118

>>58377092
Nope.

>> No.58377126

>>58377092
Or it'll be better than Intel.

>> No.58377134

>>58377092
Perhaps

They're going to have a k i3 and pentiums with SMT

>> No.58377142

>>58377092
No, or at least very doubtful.

>>58377126
Even is Zen is everything you hope it will be intel prices may not change.

>> No.58377159

>>58377142
I could see this happening

They'll only offer deals to OEMs

>> No.58377195

> not caring until a product is released
> /G/
Pick one

>> No.58377202

>>58376265
I remember this. Toms hardware in the past was caught taking bribes from Intel to produce articles like this.

In that video they used newer Intel boards with thermal shutdown protection and used older AMD boards that didn't. They were pushing a narrative that AMD was inferior when they could have just used a newer AMD board with thermal protection that was available.

Toms Hardware was well known in the PC hardware community back then as being shady as fuck.

>> No.58377250

>>58377202
Back in the bad old days Tom himself from Tomshardware met a guy from the AMDZone at a press event, and gave him his room key. The AMDZone member who happened to be straight promptly went home and posted online about it.
It was pretty funny.

>> No.58377281

Well AMD doesn't really know how to price their stuff anyway. They price unlocked variants of their processors for the same as locked ones IIRC. They probably just said "fuck it" and just decided one variant is enough.

>> No.58377317

>>58376636
> Does this mean I should hold off from purchasing a laptop?

>> No.58377353

>>58377317
You buy what you need, anon. I still have my shitty Toshiba laptop from 5 years ago, so I shouldn't be one to judge you.

>> No.58377356

>>58377281
Pricing is pretty much the only thing left on the table. Everyone under NDA discussing it on twitter is pretty optimistic, they might not be that expensive.

>> No.58377404

What's a realistic timeframe that I will be able to buy Zen+mainboard without having to camp out and hope for a lucky click to get one in my cart before they are gone?

>> No.58377452

>>58377404
They're being firm about Q1 release, and word is that volume is high. Mobo vendors are really hype about it as well, so they're expecting sales to be good.
Q1 could mean anytime between today and the end of March. No exact date as of yet.

>> No.58377477

How likely is for a devastating bug to surface once Ryzen launches?

>> No.58377487

>>58377477
No way AMD is going to make Phenom 2.0.

>> No.58377499

>>58377404
>>58377452
Mobo manufacturares in CES say they will release before March 31

>> No.58377509

>>58377477
Only in edge cases.
You won't get another TLB or FDIV bug if that's what you're thinking.

Maybe 'minor' bugs like USB and TSX like Intel had over the last few years

>> No.58377512

>>58377404
Launch day after the real benchmarks come out.

>> No.58377513

>>58377477
There were some known bugs in the A0 stepping engineering samples, CanardPC talked about them after publishing their review. They had to do with AMD's implementation of SMT and the uop cache.
Though even they are saying that these are gone in later steppings, so things are looking good.

>> No.58377562

>same shit but slower from a different company
>shit tier amd memory controller

How is anyone legitimately able to get excited?

>> No.58377574

>>58377477
Huge bugs like the Phenom TLB one aren't very common. Those kinds of things are usually shook out by release time. It's possible there could be a small bug that requires a fairly decent amount of effort to trigger though.

>> No.58377581

>>58377562
[citation needed]

>> No.58377590

>>58377581
Prove to me AMD isn't shit

That's right you can't newfriend

>> No.58377592

>>58377562
AMD has been using a 3rd party memory controller for years.
The one in Zen is apparently designed in house. Its touted as a major improvement along with the cache subsystem.
Don't spread FUD.

>> No.58377599

>>58377590
burden of proof is on you, friendo

>> No.58377619

>>58377590
AMD's last major CPU lineup was a massive failure, but I asked you first, anon.

>> No.58377643

You know, a week ago I believed Nvidiots and Macfags were the biggest morons on this board, now It seems I have to reconsider the Intel ones as well.

>> No.58377676

So I'm guessing these won't work on am3+ boards like they have in the past?

>> No.58377685

>>58377676
A SOCKET A YEAR KEEPS THE AMDGOYS IN FEAR

>> No.58377699

>>58377676
Nope, new socket. AM4.
The X370 AM4 boards are looking absolutely gorgeous though.

>> No.58377720

>>58377699
>gorgeaous
>mobo basically empty

Lol, Intel is higher quality, it even shows in the motherboards.

>> No.58377723

>>58377676
No.

>>58377685
AM3+ is from 2011, shitposter-kun.

>> No.58377732
File: 152 KB, 1441x977, 1363840695132.jpg [View same] [iqdb] [saucenao] [google] [report]
58377732

>>58377643
You sure about that?

Competition drives prices down, and performance up, its a win win.

Wishing ill on either side is just fucking yourself.

>>58377676
>>58377699
AM4 looks alot like AM3, but its diffrent. New boards do look good.

>> No.58377733

>>58377592
Too many delays. Intel is already on their fourth generation ddr4 controller. Given the circumstances I wouldn't be surprised if it only supports the bare minimum ddr4 speeds.

AMD poojeets will be out and about on how ddr4 speed doesn't matter and call everybody gaymen faggots while touting the few encoding benchmarks it wins at. Power consumption doesn't matter anymore either.

>> No.58377737

>>58377720
Could you please stop shitposting Zen threads. At least for a day. This is getting stale.

>> No.58377744

>>58377720
>Cluttered = better

>> No.58377755

>>58377723
>no new CPU since 2011

HOW CAN ANYONE TAKE THIS MEME MANUFACTURER SERIOUSLY

>> No.58377762
File: 572 KB, 600x580, goldface.png [View same] [iqdb] [saucenao] [google] [report]
58377762

>>58377733
>Power consumption doesn't matter anymore either.

>> No.58377771

>>58377732
>Wishing ill on either side is just fucking yourself.
He's not wishing ill anyone. He's just amazed at the amount of salt.

>>58377733
>I see into the future and my opinions are facts

>> No.58377779

>>58377732
AMD isn't competition.

>new boards look good
Empty like AMD's promises, no features.

>> No.58377780

>>58377733
Some boards shown at CES showed support for DDR4-4000 actually. Getting 4266 kits to boot won't be an issue.
And I'd say that solid performance in a real worth workload at low power consumption definitely matters.
Nice try though.

>> No.58377802

>>58377771
>I see into the future and victory belongs to AMD

Who *really* makes more sense here?

>> No.58377803

>>58377779
What features would you prefer, anon?

>> No.58377804

>>58377755
>socket equals cpu

THIS IS WHAT INTEL GOYS ACTUALLY BELIEVE

>> No.58377805

>>58377676
Obviously not, the socket layout is completely different because AM3+ CPUs did not have PCI-E lanes on the CPU.

>> No.58377816

>>58377803
no thunderbolt no 40+ PCIe lanes no USB 3.1 gen2, no premium VRM and ESD shielding, no DDR4 quad channel, etc

>> No.58377817

Reminder to report shitposts to improve board quality.
No reason to tolerate 9gag level bullshit here.

>> No.58377823

>>58377802
I

>> No.58377833

>>58377802
>Implying he thinks AMD will win

He's just mocking anon's futuresense abilities.

>> No.58377834

>>58377780
>DDR4-4000 actually. Getting 4266 kits to boot won't be an issue.

lol?

>> No.58377841

>>58374376
Why do all those motherboards look so shit?

>> No.58377842

>>58377817
/gee/ is in dire need of quality jannies.

>> No.58377859

>>58377816
I have literally all of that except USB 3.1 and it's pointless since my CPU takes like 70W at full blast. Okay, maybe not 40+ PCI-E lanes. I have 40 exactly.

>> No.58377865

>>58377841
>tripfag talking shit about anything at all

>> No.58377878
File: 108 KB, 2048x1152, 1363123465987.jpg [View same] [iqdb] [saucenao] [google] [report]
58377878

>>58377771
>You know, a week ago I believed Nvidiots and Macfags were the biggest morons on this board, now It seems I have to reconsider the Intel ones as well.
Yeah a completely passive post...

Anyone who buys hardware now has to become on online crusader defending there purchase because they have the self confidence of a teenage girl.

>> No.58377888

>>58377841
>Tripfag

>> No.58377934

>>58377816
AM4 boards have native USB 3.1 gen2

>> No.58377936

>>58377878
You only see amd users justifying their purchase. I'm pretty sure amd includes a flash drive full of cherry picked jpegs too as an amd starter pack

>> No.58377949

AM4 supports ddr6

>> No.58378015

>>58377934
Its not native to the CPU, but is provided by the supplemental chipset.

X370 has two 3.1 G2 ports, six 3.1 G1 ports, and six 2.0 ports.
The CPU itself supports four 3.1 G1 ports, and four 2.0 ports.

>> No.58378023

>/g/

>> No.58378032

>>58378015
Is there something wrong with that, anon?

Sorry if i'm out of the loop

>> No.58378035

>>58377936
You only see intel users justifying their purchase. I'm pretty sure intel includes a flash drive full of cherry picked jpegs too as an intel starter pack

>> No.58378047

>>58378015
>Its not native to the CPU, but is provided by the supplemental chipset.
Native to the platform. Intel boards gets Gen2 support by external ASmedia chips

>> No.58378050

>>58378032
Not at all, just providing info since facts medicate my raging autism.
The CPU+chipsets actually provide tons of IO. They're not really lacking in anything.

>> No.58378066

>>58378015
Isn't the Soutbridge integraded?
At that point the latency will be negligible.

>> No.58378085

>>58378066
>Isn't the Soutbridge integraded?
Yep. Summit Ridge/Ryzen is an SoC.

>> No.58378118

>>58378085
So there's no difference unlike with Intel 'running some PCIe lanes off the chipset' and running some off the CPU

So no matter your configuration you'll still get top latency even if you put in a bunch of PCIe cards and M2 drives.

>> No.58378133

>>58378066
AMD AM4 platform has a "PCH" (I don't know AMD terminology for it) with most things integrated into the CPU and the PCH providing some additional I/O. It's very similar to Intel.

>> No.58378265

>>58378118
10GBE, x4 M2 and a x16 GPU is all I need to run off the CPU lanes, so 26 should be enough.

>> No.58378601

>>58376299
wtf are they doing

>> No.58378627

>>58376503
>tech always moves forward just make your purchase.
Make me purchase something other than my 2500K

You cant.

>> No.58378629

>>58378601
Your mom.

>> No.58378966

>>58376918
unless intel refuses to let you oc on any motherboard, you still need good power delivery with them too.

>>58377026
its the possibility of a fairly cheap, fairly powerful 8 core, and worse come to worse price wise, a cheaper 6 or 4 core should be in the i7 to i3 range.

If amd just pushes a minimum 8 core out for 350~ 6 core for 250 and 4 core for 150, which they could easily do at proffit, it brings to question why bother with intel for anything 4 core and down, and depending on what aspects amd fails at, possibly 6 core and down, with a serious 'do I REALLY need that much power for the 8 core'

I honestly don't think amd will sell minimum 8 core for over 500$ for a few reasons, one being their marketing, the way they talk about it and show it off, and if they price the 8 core within 200$ of intel, most people will still go intel because its the proven quality for the last near decade. amd needs to do something big on the consumer side to get people back and making everyone question 'why even consider intel' would do that, if they go into a pricing war with intel, where intel decided to match amds price if they even decided to lower it at all, then its just a slow trickle down where no one will pick amd at any price, this is a chance to make a fairly large profit, get people on an amd platform, and if they have a real jump in quality with zen+ they have people already invested in the platform, so they have a second buy, and if they keep the platform long enough a third as well. You are not going to make sales if you price base at 500$ as intel's 6 core is cheaper and will be 100% of the power 100% of the time, whereas amd only has 80% of the power (from their mouth, they made a cpu that will compete in 80% of the tasks given to it)

we are counting down a 3 month period of time when a complete shift in what we call mainstream could happen, an absolutely horrible time to make a major investment in a computer unless you have to.

>> No.58379748

>>58374590
You can get a cheaper motherboard with AM4 and not have overclocking as well.

It's just that if you want to get overclocking on your 500$ AM4 CPU, you don't need to buy a new CPU, you buy a likely less than 200$ Motherboard that supports it.

>> No.58379774

>>58374866
Papermaster clarified that it wouldn't be the end of Q1 like many other launches have been.

Expect mid to late Feb to early March launch.

>> No.58379834

>>58375932
32c opterons at 3Ghz should be pretty reasonable.

AMD is about to chew a huge hole into Intel's marketshare across the board.

>> No.58380040

>>58374284
>unlocked CPU
>locked chipsets.

>> No.58380074

>>58374376
>currytech
This shit needs to be autosage

>> No.58380268

>>58379834
It'll be really interesting to see its voltage/clock curves. Back when the only known figures were 2.8ghz/3ghz for an early engineering sample Ian Curtress who writes for Anandtech guessed that the Naples chips would be clocked at only 2ghz.
It comes down to that curve, and inside of 95w, with about 15w being taken by the uncore, Zen is consuming about 10w per core to hit 3.6ghz in these latest chips. Excavator in Carrizo needs 10w to hit 3.15ghz, but cutting the power in half to 5w still gets you 2.45ghz.
50% reduction in power, over 75% of the clock speed. If Zen had the same scaling then 2.7ghz or higher should be doable.
Enterprise die are naturally the highest binned ones, so they'll have a slight advantage in their favor for clocks and voltage too.

At present the biggest baddest enterprise CPU from intel is the Xeon E7 8890v4. A 24c/48t monolithic die behemoth clocked at just 2.2ghz. With similar IPC a 500mhz clock advantage would give Naples the edge in tons of metrics, not to mention its 8 memory channels.

Things might really change for AMD. Exciting times.

>> No.58380346

>>58380268
>AMDs entire SR/Naples lineup is 1 (one) SKU.
Talk about cost efficiency.

>> No.58380511

>>58380346
They've done it that way for years, just never on this scale. Of course they haven't had such a competitive architecture in years either.

>> No.58380568

>>58380511
Also unified platform for anything Zen-related. Fuck me, I would be able to upgrade from 4c/8t to 8c/16t Zen without changing my mobo.

>> No.58380612

>>58375790

My 290x disagrees. While I love the card. Great performance and I got it for like 175$ two years ago (dirt cheap in norway)

It heats up my room very fast. I don't need an extra heater in the winter. And I'm not even joking.

It performs close to a 980 in t he games I play so I'm happy

>> No.58380637

>>58380612
You got a bad binned one I guess.

>> No.58380664

>>58376240

When the 780ti was released it was a bit better than the 290x

The 290x shits on a 780ti now and is on pat with a 980. The 390x beats the 980 in almost all games now.

>> No.58380698

>>58375790

>AMD house fires meme

the r9 2xx series, 295x2 and FX-9590 are living proof.

>> No.58380724

>>58375140
Because they've sold zero units so far :^)

>> No.58380745

>>58380724
Well memed, my friend.

>> No.58380830

>>58380724

Kekd

>> No.58381160

>>58380612
Not bloody likely. That is maximum 275W part. That is approx 100W more then a competetive intel product.

100W is what a decent light bulb would put out years ago. Typical heater is rated to kilowatts.

I own a 290 and while the exhaust is hot, it barely is noticable in about 1-2 meters away from the PC.

>> No.58381347

>>58375519
>Intel Designed Thermal Final Solution

>> No.58381386

>>58375712
This. First posters of Intel shills in this thread should read this

>> No.58381422

>>58376559
>damage control

>> No.58381755

>>58374647
There not even fucking released yet. They were supposed to come out last year. It's been a terrible year

>> No.58381791

>>58381755
>They were supposed to come out last year.
No, launch for Summit Ridge was never 2016. AMD only stated limited availability for late 2016, and they met that with OEM sampling.

Nice FUD though, shill.

>> No.58381854

>>58380346
that's the number 1 reason aside from die size I believe the 350$ price point for a bare minimum 8 core, its literally the worst binned chip they are going to put out, going to a 500$ one or whatever their step up price is would probably net you something closer to intel for overclocking.

>> No.58381893

>>58381160
to be fair, all gpus, regardless of brand, heat up a closed room fairly quickly, amd just gets you there slightly faster than nvidia.

Had this argument so many times with my brother over his water cooling setup "My cpu and gpu are so much colder now that they aren't air cooled, my room won't get hot at all" and the ass refuses to believe his room is a sauna because of his gameing.

>> No.58381907

thank you based AMD

>> No.58381910

>>58380698
still have yet to see amd hardware throw sparks on camera (when you are not trying to destroy the part) , but have multiple nvidia ones throwing sparks, thankfully the cards were all brand new, so dust didnt accumulate.

>> No.58381930

>>58381854
Intel 8-cores OC like hot shit.

>> No.58381938

>>58381893
Tell him that conservation of energy is king. Unless his radiator is in the other room, then all that heat is dumped into his room anyways.

>> No.58382005

Should I sell my 5820k with mobo and wait for Ryzen? I am kind of an AMD fanboy...

>> No.58382032

>>58382005
No.

>> No.58382083
File: 378 KB, 1440x2560, 20170108_025308.png [View same] [iqdb] [saucenao] [google] [report]
58382083

>amd posters

Nuke them all.

>> No.58382276
File: 103 KB, 700x400, 55FPSTitanXpisNotAcceptable.jpg [View same] [iqdb] [saucenao] [google] [report]
58382276

Sorry but linuscucktips demo on a marketing build is running BF1 at 54 FPS on a Titan XP jesus AMD made a workstation cpu yet focused on a market it completely fails at AMD IS FINISHED

>> No.58382492

>>58382083
>install ubuntu

>> No.58382571

>>58374628
AMD hasn't been relevant to anything CPU related since phenom II.

>> No.58382928

OP here, seems like Intel shills have given up. Let this be a reminder that Intel is literally shaking right now

>> No.58383146

https://translate.googleusercontent.com/translate_c?depth=2&hl=en&ie=UTF8&nv=1&prev=_t&rurl=translate.google.com&sp=nmt1&tl=en&u=http://pc.watch.impress.co.jp/docs/column/kaigai/1037983.html&usg=ALkJrhjgXsNBjn3hUq3S4m4S5rad61i_Aw just something about the new cache subsystem ryzen is using. Dresdenboy posted it on his twitter

>> No.58383389

>>58374284
weren't all of Bulldozers overclockeable as well?

That did go well for them.

>> No.58383412

>>58383389
Different arch.

>> No.58383478

Intel Core 2 Quad master race.

Enjoy being cucks.

>> No.58383800

>>58377841
I like the edgy swirly gear one on the left.

>> No.58384108

>>58374284
No no no, Intel was finished again last week. This week, some other company is finished. Proper protocols must be followed.

>> No.58384172

>>58381930
depends on what you call shit, really even at base clocks they are more than fast enough even for gaming, but nearly every 8 core hits 4.3ghz, again, more than enough for gaming.

But when you do something that actually uses all 8 cores, even that oc really adds a fuckload.

>>58381938
I gave up. I realize that there are some things that people just can't wrap their head around unless they understand it.

my mom is that way with stars, she doesn't grasp how a star can supernova and scientists say it happened a number of years ago.

He is that way about heat, he sees the numbers are lower, so it must not be getting as hot, what he doesn't get is he is just dumping the heat faster.

>> No.58384186

>>58382571
to be fair, bulldozer and its revisions were fantastic when shit actually used all the threads, problem was next to nothing normally used made use of all those threads.

>> No.58384743

>>58374284
I wonder how much 4c/4ht zen + mobo + 2x8gb ddr4 3000mhz ram will cost. I hope under 400 euros or i will go bankrupt, i barely collected $280 so far

>> No.58384997

>>58384743
It's 4c/8t, AMD is not going to disable SMT just because.

>> No.58385009

INTEL IS DONE

SELL IT

7NM NOW

>> No.58385281

>>58376570
I care for multiple reasons. First, I do vt-d so I dedicate the discrete gpu for guest os and host os uses igpu. Secondly, while this isnt exactly to the point, I build workstations which doesnt include a discrete gpu. Some clients who do some heavy computing simulations have no usd for discrete gpu.

>> No.58385369

>>58374348
>t. Intel Marketing Team

>> No.58385428

>>58377834
already announced

>> No.58385533

>>58377780
Was there any huge difference with using DDR4-4000 vs DDR4-2133 or its a meme? I hated Intel for locking 2133 on B150 boards

>> No.58385575

>>58385533
there is, and pretty big one
DDR4 is finally worth it

>> No.58385695

>>58385369
Then make a Zen thread with OP that is not shitposting magnet.

>> No.58386066

>>58376997
hahaha
> high tech car
> powered by a weak cpu with a 250W TDP

yeah, huge chance bro

>> No.58386101

>>58385533
B250 already supports DDR4-2400 and anything higher is pointless because of dimishing returns and the fact that high speed RAM without shit latencies cost a shitload.

>> No.58386111

>>58386066
hahaha
>high tech car
>powered by a locked down piece of shit cpu that can't even be overclocked without paying musk $1000

yeah, huge chance bro

wait....

>> No.58386137

>>58374590
my cheap ass am3+ mobo supports overclocking...and it was a fucking 25 dollar bundles deal with the chip.

what fucking motherboard is so cheap it have 0 oc support? a prebuilt maybe?

>> No.58386561

>>58374866
Wtf I hate benchlife.info now.

>> No.58387083

>>58374866
We'll look into that. If AMD delivers then good for them, time to shake up the industry.

>> No.58387868
File: 120 KB, 720x960, d25d4d0e19cbe8db3528848f7175f0ab4998a8daa9d7d327f8b2a7dab373f280.jpg [View same] [iqdb] [saucenao] [google] [report]
58387868

Kek

>> No.58388318
File: 434 KB, 1920x1080, AYYMD-on-suicide-watch.jpg [View same] [iqdb] [saucenao] [google] [report]
58388318

>>58387868

>> No.58388918
File: 80 KB, 680x752, aaf.jpg [View same] [iqdb] [saucenao] [google] [report]
58388918

>>58374284
>>58374376
>Khalid Moammer
>wccftech

Literally a poo website that pull shit out from there asses claiming baseless rumors and retarded "predictions". All they write pretty much taken from other reputable review site and regurgitate it and they add their own "opinions". The website is owned a by fucking shit skinned nigger creating articles to feed his obese Indian children. Don't bother, the site is a literal AMD shill.

>> No.58389031

>>58382083
Jesus Christ, the fucking normies ruined this site, makes the summerfags look better.

>> No.58390444

>>58388318
DELET THIS RIGHT NOW

>> No.58390531

>>58376623
>http://www.pcworld.com/article/3155109/computers/new-amd-ryzen-details-revealed-overclocking-crossfire-lineup-info-and-more.html

Maybe not high end but still dependent on the chipset

>> No.58390571

>>58390531
So budget boards with weak VRM don't run into throttling and power delivery issues.
Don't be retarded.

The high end and mainstream chipsets support overclocking.

>> No.58390583
File: 366 KB, 1365x762, MuhIPC.png [View same] [iqdb] [saucenao] [google] [report]
58390583

>>58388318
>>58390444

Fine since you're trips. one second

>> No.58390927
File: 224 KB, 778x960, GDyOS.png [View same] [iqdb] [saucenao] [google] [report]
58390927

>>58374284
but how many cores does it have

>> No.58390999
File: 36 KB, 549x477, IntelCES2017.jpg [View same] [iqdb] [saucenao] [google] [report]
58390999

>>58390927

>> No.58391060

>>58390927
When I mean rekt. Linus has bend the knee.

https://www.youtube.com/watch?v=vMfNz2SXVLk

like that one anon said "If Amd were to get competitive in the CPU market, it would make Intel CPUs cheaper."

Competition is good for the consumers

>> No.58391063

>>58375911
>So if AMD being housefire is a meme, so is the GTX 480 being a housefire
GTX 480 being housefire isn't about the TDP, but the fact that it hits 100C under load.

>> No.58391124

>>58376945
>Phenom X4
>GTX 460
Jesus Christ nigga. MIght as well wait some more for Zen. On the GPU, literally anything will do, even the shitty RX 460 will do. I'd get a RX 480 and pray it's from the better binned ones. That or maybe wait until RX570 for them to be incredibly binned.

>> No.58391136

>>58374284
Presumably because they will need to be underclocked to run reliably. Wait till you see the ADSL style "up to" on the packaging.

>> No.58391157

>>58389031
Fuck off, normalfag.

>> No.58391374
File: 314 KB, 478x485, ayymdhousefires.png [View same] [iqdb] [saucenao] [google] [report]
58391374

Let this thread end with this picture

>> No.58391575
File: 218 KB, 1000x505, Vega architecture.png [View same] [iqdb] [saucenao] [google] [report]
58391575

>>58391374
No.

>> No.58391740

>>58391575
Looks good on paper. Where's the so-called Livestream again?

>> No.58392637
File: 1.32 MB, 720x1280, AMD ZEN.webm [View same] [iqdb] [saucenao] [google] [report]
58392637

ZEN WILL BE GREAT
STOP LAUGHING

>> No.58392730

>>58381347
Underated post

>> No.58392957

>>58374427
>gee let buy this unlocked cpu
>no VT-D
get cucked

>> No.58393029

>>58380664
I think you meant to say in none.

>>
Name (leave empty)
Comment (leave empty)
Name
E-mail
Subject
Comment
Password [?]Password used for file deletion.
Captcha
Action