http://www.digitimes.com/news/a2016 0621PD205.html IT'

HOLY SHIT THEY CAN'T EVEN GET USB 3 TO WORK RIGHT

AMD CONFIRMED INCOMPETENT AT CHIP DESIGN

>>55199945
We already knew that

>>55199909
They should let third parties make chipsets for them. Nvidia used to make chipsets for AMD, ULi (now part of Nvidia) did also. Then there was everyones favorite VIA back in the day. Like why not?

Amd can't support thunderbolt because that's intel's technology right?

>>55199909
>indiatimes

>>55200010
So close
>>55200000

>>55199989
They should let Nvidia set up their vram.

>>55200049
They should let Nvidia just do everything for them

>>55200049
>3.5gb
Nope

File: 1452324829767.jpg (52KB, 749x444px) Image search: [Google]

52KB, 749x444px

at the very least it isn't being kept secret and they admit problems .

https://en.wikipedia.org/wiki/Pentium_FDIV_bug

How much would this effect enterprise? Because from what I understand, AMD is targeting enterprise with Zen because that's where all the money is and that's also where they've lost the most to Intel. If true, could this be a deal breaker for enterprises looking to use Zen?

Is this really legit? It seems kinda late in the game for them to just be noticing something like this.

zen on track? not likely considered zen had still not taped out in december of last year. it's gonna be at least 15 months from then until we see a real, working zen cpu available to purchase for consumers. we won't see anything other than some partially validated samples being given to reviewers as a paper launch before that.

>>55199989
>They should let third parties make chipsets for them
They already do. This is designed by ASMedia, who have previously had problems with SATA and USB performance. They're calling this a market rumour and it's said that past performance is not a reliable guide to future performance but it's the only one we have.

>>55201829
:^)

Correct me if I'm wrong, but doesn't the article say that it will only incease the motherboard cost two to five U.S. dollars?

>>55201829
Stop spreading this nonsense. It was taped out some time ago, samples have already been sent to close partners and samples are being sent to other partners next month. They are ramping production soon and will be making Zen available Q4.

I know this because I actually pay attention to investor calls because I own stock.

>>55202034
will zen be good

>>55201770
Nobody gives a shit about USB on servers. The only time you ever might possibly even need it is if you need to direct console in because shit is FUBAR, and even then you don't need USB 3.1

It won't affect enterprise at all.

>>55202040
Should be. Looks to be slightly ahead of haswell in ipc since they exceeded their 40% ipc uplift estimates. So slightly better than haswell, SMT and 8 cores. Zen looks pretty good, should help them take back some of the server market, and once raven ridge launches next year large orders for businesses needing "workstations" should roll in. Combine that with the business notebook market and it should be a decent money maker. Zen is targeting enterprise, largely in the server market, where it will meet 80% of that markets need.

>>55202091
neat .
i'm still heating my room with an 8320

>>55202132
My 8350 is going to get retired to my server and undervolted. Shit is only going to serve web pages and mumble anyway. I'll throw a couple of GPU (maybe the 480) into the matx build so I can offload compute to it for fairly cheap.

This latency issue is pretty well known for usb 3.1


Also like i care intel and amd is best combo prove me wrong

>>55199909
ah digi times the same site that said that polaris will launch on october because it had thermal issues

good on them being right on track with videocardz/wccft/hardocp on the worst sites of 2016

>>55202157
the 8 core fx gets a lot of flak but it has served me very well

>>55202216
what sites do you recommend anon

>>55202216
I think videocardz is actually improving. I would not say it's good, but it's better than before at least.

>>55202232
none desu 90% of them are bias or slightly bias towards nvidia or amd or intel no one is trully neutral...

i guess the best thing to do is to wait for the massive review sites to do their work and exclude the ones that show quite a lot of difference between them
(gamernexus will be the one pcper will be the other some youtuber for sure will show stupid results too)

>>55199909
>This probably means that USB 3.1 will be limited to rear i/o

Okay. So what?

>>55201829
Kill yourself intelshill.

Intel had chipset issues every fucking generation since Sandy Bridge and nobody cared.

>>55202091
40% ipc increase compared to Bulldozer, still way bellow haswell

>>55203383
30% increase over excavator is significantly more than 40% over bulldozer.

>>55201770
>>55201988
>>55199945
>>55202189
>>55202923
>>55199909

Intel uses a third-party chip on skylake for 3.1 support. Only kaby lake will bring native 3.1 support.

>Commenting on the news, AMD said it is pleased that Zen is on track and will not comment on customer specific board-level solutions., while ASMedia clarified that this is purely a market rumor and its product's signal, stability and compatibility have all passed certification.

Did anyone actually read the article?

>>55203213
at least my x97 is better than my old x58 i used to use but their both still shit.

amd chipsets usually arnt as bad but have their own issues.

>>55202034
>stop saying facts that i don't like

even intel can't tapeout a design and have it validated and ready in less than 12 months you idiot. every AMD architecture since k8 has taken at least 15 months from tapeout to availabiltiy as well.

zen won't be here before q1 2017, stop being a delusional shill.

>>55207348
Su said herself that it'll be out by the end of the year.

>>55207735

why do you care what a marketing hack who drove a billion dollar company into the ground says on a stage?

the fact is that no CPU has ever gone from tapeout to availability in less than 12 months...since zen didn't take out until well after december '15, this means we can only see it in q1 or q2 2017 at the earliest.

>>55207787
It's okay anon, I'd shill for Intel as well if they asked, they pay well.

>"which will increase motherboard players' costs by US$2-5"
wow it's fucking nothing

>>55207881
2$ for them is 10$ for us

AMD IS FINISHED

>>55207892
Source: my ass

File: 1466432175151.jpg (42KB, 604x604px) Image search: [Google]

42KB, 604x604px

>>55206138
HOLY SHIT THEY CAN'T EVEN GET USB 3 TO WORK RIGHT!

INTEL CONFIRMED INCOMPETENT AT CHIP DESIGN!

>>55199909
Will be fixed in the C2 stepping a few months after release.
Who the hell cares.

>>55207940

because it costs millions of dollars to respin and deprives early adopters of functioning CPUs

>>55207961
Most CPUs go through at least one stepping from release, I don't see the problem.
This one will just be earlier.

>>55207961
>deprives early adopters of functioning CPUs
Yeah, it deprives them of USB 3.1 from the CPU which they'll still get from third party controllers on the mobo, costing 5$ more

Wow.

>>55208029
>pre release price
>Eastern Yurop

File: 1070 is cheap they said.jpg (41KB, 496x687px) Image search: [Google]

41KB, 496x687px

>>55208029
It's not any better on Nvidia's side

AMDPOORFAGS ON SUICIDE WATCH!!!

>>55208029
>slovaks on /gee/
what

>>55208052
I'd say avoid founders editions, but outside of those this is the cheapest I've found

https://www.pcdiga.com/2/21572/MSI-GeForce-GTX-1070-GAMING-X-8G-912-V330-001

>>55206138
>Intel uses a third-party chip on skylake for 3.1 support.

False: https://encrypted.google.com/#q=intel+Alpine+Ridge

>>55199909
> or even an independent USB 3.1 IC
OH NOES! Nobody will buy it then!

>>55199909
The stock is down 20%. Oh shit poo-laris should have just stayed in the loo

File: totally cheaper than rx480.png (428KB, 1423x954px) Image search: [Google]

428KB, 1423x954px

>>55209659

>>55209659
So 310 USD, no "beast mode" cards confirmed

>>55207787
> a marketing hack who drove a billion dollar company into the ground
That wasn't her. Some white male did it.

File: article_2638474_1E2ED4EB00000578_615_306x423.jpg (29KB, 306x423px) Image search: [Google]

29KB, 306x423px

>>55209553
Wether or not they use a third-party, or design their own is immaterial to the discussion. Intel does not support native USB 3.1 until kaby lake.

>>55210176

how exactly is it not 'native support'? the controller is literally on the reference PCH that intel sells to their partners.

>>55207787
>>55207735
>>55207348

It WAS coming in Q4 2016.

http://www.game-debate.com/news/20457/both-intel-kaby-lake-and-amd-zen-cpus-may-not-be-ready-for-launch-until-2017

Both Kaby Lake and Zen have been delayed.

>>55199909

Bullshit, OP.

>>55210265
The fact that it needs a controller should clue you in.

>>55210333
>It WAS coming in Q4 2016.

yes, amd wanted to bring it out in q4 after they delayed it from q1 originally. the extra time they took fixing shitty pooinloo designs delayed it even further now too, sadly now zen is gonna be out around the same time as skylake-e and cannonlake

>>55210502

>The fact that it needs a controller should clue you in.

you always need a controller. do you think the CPU does something magic to make USB work? the only actual difference is that they put the controller on the PCH instead of the CPU, probably so they can update it without respinning all of their cpu designs.

>>55199945
usb 3.1 that is

>which will increase motherboard players' costs by US$2-5
>$2-5
>AMD Zen chipset designs have already been finalized and will begin shipments at the end of the third quarter and start mass production in the fourth.


more fearmongering over non issues i see

>>55202028
Plus profit ;)

>>55202028
>Wanting to get cucked out of 2 dollars
There is only a limited amount of wealth in the entire world. Every dollar adds up, anon. Don't take your money for granted, no matter how little it may seem now. 20 years from now you're going to wish you held onto that precious money.

>>55210882
>limited

not when the government can freely print more.

also saving is a surefire way to lose value. you should be investing at the very least to beat inflation.

>>55207892
wow it's fucking nothing

>>55199989
Have you ever used one of those Nvidia or Via chipsets? They were all godawful.

>>55210952
4u

>>55211058
is an hour of min wage make or break for you?

If so get a fucking job you lazy NEET

>>55210952
Shiiiiiiiit, if everyone in ITT gave me 2 buckaroos, I could put it towards getting AMD's gpu which would mean you would have reinvested your money back into AMD. Sounds like a good deal imo, everyone wins, nothing of value was lost.

>>55210512

>fixing pooinloo designs

Jim Keller is not a Pajeet

Asm 1142

This is what is on my skylake mobo.

It's not what ever the fuck you were going on about earlier.

Again: you are a fucking idiot

I would post the website, but it's apparently spam.

>>55210553

Asmedia
Dot
Com
tw/eng/e_show_products.php?item=158&cate_index=98

>>55210553
But they re pin all their shit


A socket change a year keeps the goyim in fear

>>55199909
>IT'S OVER
Say that to my wallet

>>55207348
Taoeout was long ago, AMD has shown working ES
Also on track could mean anything, AMD hasn't set a precise date for Zeb

>>55209844
Actually it was a spic, though he was around in the times when AMD wrecked Intel

>>55202216
Videocardz and wccft have been quite right regarding Polaris
>>55210512
It was never scheduled for Q1 fucking shill
Also, there no pooinloos at the CPU division

>>55202216
I hate currytech so much, but I don't know where else to go to get hardware rumors. Fortunately after Polaris, Vega, Pascal, and Zen, I can forget about them for a while

>>55199909
>"Commenting on the news, AMD said it is pleased that Zen is on track and will not comment on customer specific board-level solutions., while ASMedia clarified that this is purely a market rumor and its product's signal, stability and compatibility have all passed certification.

AMD Zen chipset designs have already been finalized..."

Can you fuckers not read?

>>55213421
DAMAGE CONTROL

>>55209698
Britain has high tax on electronic shit :(

who cares about usb 3.1 anything over 2.0 is redundant

>>55210265
alpine ridge is the thunderbolt controller dimwit

>>55211410

he's worse than a pajeet. he makes a design an then leaves before it's validated so that the pajeets have to go bug hunting in his million lines of shitty verilog

more likely than not zen's delay is probably directly attributable to him not wanting to do the 'menial' validation work and instead demanding to work on zen+ before he jumped off the sinking ship