Will he be fired if Vega and Polaris flops?

>>55406154
I don't get why they used the 6 pin. Would've been so much better to either use a single 8 pin and include a 6 to 8 adaptor or use 2x6pin.
Absolute madness.

>>55406189
I want to see how rx 480 with a 8 pin power connector will fare. There is a Sapphire Nitro model of the rx 480 but it won't be out for 2-3 weeks. I think with proper power it should be able to do 1.7-1.8Ghz.

File: 14-23-20-AMD-Radeon-RX-480-Cooler.jpg (272KB, 1200x750px) Image search: [Google]

272KB, 1200x750px

>>55406239
I think with a proper 8 pin and cooling solution it'll be within spitting distance of the gtx980. Right now it's only a little ahead of the 970.

>>55406239
More like 1.4-1.6Ghz

>>55406189
More pins means more VRMs

File: 8a.jpg (78KB, 800x662px) Image search: [Google]

78KB, 800x662px

A 6-pin can provide the same amount of power that an 8-pin. The only difference between the two are two extra grounds on the 8-pin.
Where AMD fucked up is in the way the 480 draws its power. The card limits itself to drawing 75W over the 6-pin and gets any extra power it needs from the PCIe lane when it should be the opposite.

Vendor cards are going to fix all this and allow additional overclocking headroom, unlocking the RX 480's true potential.

>>55406295
>it will be fine, only reference cards fry your motherboard

Hilarious

File: CmNlxTVWAAAxYnm.jpg (127KB, 1200x676px) Image search: [Google]

127KB, 1200x676px

>>55406295
They'll also stop it from being a hoverboard

>>55406154
>extremely limited quantities of HBM2 being manufactured
>wasting it on Vega

Such a shame.

>>55406295
>A 6-pin can provide the same amount of power that an 8-pin.
That's not how it works. 6 pins is 75W and 8 pins is 150W.

>>55406308
It'll be fine, only reference cards are a extra $100

lol?

File: ayymddv3.jpg (695KB, 1920x1080px) Image search: [Google]

695KB, 1920x1080px

>>55406239
>1.7-1.8Ghz

in your dreams, most rx 480s won't even achieve 1.5ghz with 2x 8pins

>>55406295
>A 6-pin can provide the same amount of power that an 8-pin.

you are biggest idiot i ever seen

>>55406351
Actually it is, you can draw as much as you want. But less ground pins means higher stress on the psu

>>55406487
>in your dreams, most rx 480s won't even achieve 1.5ghz with 2x 8pins
Why? GTX1080 does 2Ghz on 16nm, why won't rx 480 be able to do 1.7Ghz on 14nm? I know AMD likes to fuck things up but 1.7Ghz should be possible.

>>55406351
That is exactly how it works. An 8-pin doesn't have any extra connectors that allows it draw extra power, it just has two more grounds than a 6-pin. A 6-pin can safely deliver 150W, the only reason it doesn't is the PCIe specification.
The ONLY reason the RX480 draws more power than it should from the PCIe lane is either a hardware (hopefully not) or software issue causing the card to draw any power it needs beyond 150W from the PCIe lane, potentially damaging older or cheaper motherboards.
It's a shitty, almost inexcusable mistake from AMD that's already done serious damage to the RX 480 from a PR point of view.

>>55406557
Different architectures. Smaller node doesn't necessarily mean higher clocks. Aftermarket 480s will probably do 1.5GHz at absolute best. Don't try to buy into the hype, you're just setting yourself up for disappointment.

>>55406557
because hes a shill

>>55406557
Nvidia can reach 2Ghz on Pascal because its Maxwell but shrunk and has a SMALLER core count than Maxwell.
Less cores = less to go wrong at higher clocks, also means that there's less heat being passed along from core to core, so higher tolerance to clocking.

With more cores come slower, but more powerful overall 1:1 clocks.

>>55406511
dumbass here, how do more ground pins reduce psu stress?

>>55406616
>Nvidia can reach 2Ghz on Pascal because its Maxwell but shrunk and has a SMALLER core count than Maxwell.

Plus TSMC 16nm is better than GF 14nm LPP

AMD made a massive mistake in choosing a process designed for low powered mobile chips.

What about the 490x? You think that might be the savior AMD needs or will it just be poo in the loo?

>>55406569
>Aftermarket 480s will probably do 1.5GHz at absolute best.
We need to wait and see.

>>55406733
>TSMC
>better than GloFo
is that why TSMC is having problems with 16 nm and cant deliver anything in volume?

>>55406189
It's a clever marketing trick.

If they had put an 8 pin on the card, amdrones wouldn't be able to claim that the card is power limited. Now they can claim that the aftermarket models will have vastly better performance than the reference models.

They did the same thing with 290X, except in a slightly different way. The card ran at 95C. This brought amdrones out in droves claiming that the card is just throttling and therefore can't beat X card. Once aftermarket models come out, they'll surely perform much better.

Fool me once, shame on you; fool me twice, shame on me.

>>55406755
>trusting Charlie Demerjian

>>55406709
They don't.

File: let-down.png (609KB, 854x640px) Image search: [Google]

609KB, 854x640px

>>55406763

File: bulldozer.jpg (44KB, 575x333px) Image search: [Google]

44KB, 575x333px

>>55406763
>people still had trust in AMD after the shitdozer fiasco
>releasing your new flagship CPU that turned out to be slower than the current flagship

>>55406154
He's just running the company into the ground to make it more affordable for Intel to buy their GPU division, then he'll make a hasty exit with all his money.

File: amd dirk mayer pelagico lol.png (116KB, 576x1686px) Image search: [Google]

116KB, 576x1686px

>German site Golem.de observed identical behaviour, as did TecLab and PC Perspective. PC Perspective discovered that AMD’s card drew 7 amps over the PCI-E slot’s +12v rail, which is rated for 5.5 amps maximum. The Radeon RX 480 also overdraws the board’s solitary 6-pin power connector.

>drew 7 amps over the PCI-E slot’s +12v rail, which is rated for 5.5 amps maximum.

>drew 7 amps over the PCI-E slot’

>drew 7 amps!

7 AMPS

>>55406880
There is no way. Is it just a spike? This shit would fry any PSU and MOBO in a couple of months.

>>55406893
Yeah it's pretty bad. Not a spike, it runs like that consistently. I don't think it affects the PSU though, only the motherboard.

>>55406893

Man, my car battery spikes out 300 AMPS... it is attached to 1cm thick cables.

I'll get the fixed version for cheap while babbies flock to $800 NVIDIA cards out of fear.

>>55408242
also they are under immense pressure not to fuck it up a second time