AMD Officially Diversifies 14nm Manufacturing With Samsung

File: 1468634845764.jpg (89KB, 600x600px) Image search: [Google]

89KB, 600x600px

>>55756946
I love you Amada-chan!!!

So another manufacturer for producing even more cheap currytech shit. Got it.

Still sticking with Nvidia/Intel.

stupid slut

>>55756998

Some of us don't feel the need to pay double the price for 1/10th more performance.

>>55757065
Some of us don't want to have to pay for a new computer because a AMD card blew it up.

Some of us don't want to pay $200 for poo in loo tech that only works in 2 very specific DX12 games.

>>55756998
It will run cooler and be even better price/performance.

>>55757324
Yeah Polaris sure proved that

>>55756998
Samsung makes the best ARMs, though.

>>55757365
LMAO

>>55757381
No they don't, they just make the most amount and flood the market with their shit

>>55757448

Samsung makes the best NAND and ARM SoCs. If they could make GPUs, they would render Qualcomm obsolete.

>>55756946

Very interesting. I would hazard a guess that Samsung is more expensive than GlobalFoundries but they've negotiated capacity if Zen products really start moving off the shelves. I suppose it could instead be a sign that GloFo 14nm wafers are really underperforming and they're readying
Samsung as a last resort to save Zen/Vega. I hope it's not the latter.

>>55759339
They'd cost the same. They're running the exact same node, on the exact same machines, and running the exact same test wafers at the exact same intervals.

The agreement between Global Foundries and Samsung was that either party would step in to provide increased volume if the other needed it. Samsung really doesn't need the capacity from GloFo, but GloFo shared tons of IP with them in exchange.
The deal just gives Global Foundries the ability to quadruple part output for AMD, and they're going to need it for upcoming enterprise and OEM parts. They've already secured a few enormous contracts with a couple Chinese firms, they're supplying some GPUs to Apple, they have new semi-custom SoCs entering production, and 4 die Opteron MCMs will eat up quite a bit of that capacity for any one facility.

AMD's mid term outlook is already quite impressive. By all accounts they'll make 10billion in revenue for 2017. The next few years for them will be their biggest in a decade.

>>55759450

There's still building and labor overhead costs to consider but I concede the point--I had forgotten that GloFo is essentially leasing Samsung's 14nm.

I wonder why the Forbes analyst is so wary of GloFo's 14nm wafers if they are as interchangeable as you seem to think. (No sarcasm intended.)

>>55759556
Analyst is an ancient latin word for "man who lied himself into a position and has to keep lying or people will find out hes full of shit." Really. All of the so called analysts specializing in tech take complete shots in the dark. Broadcom was going to be the company to take down intel, ARM would own the datacenter by 2016, Windows 10 was going to be a web based OS. They're all full of it. Seeking Alpha and Forbes are both horrendous when it comes to this. Seeking Alpha in particular is basically nothing but outright stock manipulation in text form.

Fab8 in NY has a staff of Samsung engineers embedded to ensure everything mains consistent vs Samsung's 3 14nm lines.

I've read about Sammy's 14nm process not being that effective if scaled up (for GPUs) vs. mobile.

Is that true?

>>55759339

Funny enough rendering Qualcomm obsolete would render AMD also obsolete, remember ADREON is an anagram for RADEON.

>>55759692
Mobile was run on 14nm LP, GPU and x86 are run on 14nm LPP. Basically a better, higher powered sister to the mobile line.

lewd!