AMD RX480/470 PCB revealed. Apparently they share the same PCB

AMD RX480/470 PCB revealed. Apparently they share the same PCB.

Picture that with a Kodak!

OP confirmed for underage because 3DFX did the same thing with the Voodoo 3 2000 and Voodoo 3 3000 in 1998.

gtfo of my Sup Forums fgt

...

Is this the new Batmobile?

What a shit looking design.
12v power planes will have to run under the GPU, expect FUCK ALL overclocking due to noise in the PCB :)
Maybe the AIB partners won't be so fucking retarded.

They share the same PCB because 470 and 480 share the same chip.
Just like GTX1080/1070

Eat shit, tripfag.

>truth hurts
I know, but you'll see soon enough.

>I have no EE background
>I have no in depth technical knowledge
>I'm not even remotely familiar with consumer reviews

Adorable. A high OC for any previous GCN card would be around 1400mhz. If you got near 1500mhz it was miraculous due to the architectures short pipeline, and the voltage needed to hit those clocks on early leaky 28nm TSMC silicon.

The RX 480 hits at least 1200mhz stock. If you really believe that it won't be able to squeeze out another 200mhz~ then you're too retarded to help.
Have your care taker smother you in your sleep, man child.

AMD BTFO

>200mhz
And yet Maxwell overclocks were regularly 500mhz or more...

200mhz is childsplay.

Call me when 480X can OC to 2.2ghz on air.

and MHz=/=nivida MHz

You can't compare radically different architectures with different pipeline lengths like that, you tech illiterate retard.
Apple's Twister cores don't clock like stock Cortex A72 cores do either.

>still completely incapable of anything over 1300mhz without liquid cooling.

Twist it whatever way you want red hens, POOlaris is a failure waiting to happen.

>what is star ground

lol pcb noise.

AMD MHz=/=nivida MHz

>cycles per second doesn't equal cycles per second
Sure.

Actually you got right here first folks, this guy just fucken sovled it.

P must equal NP since it can't equal P!

Not sure if trolling or stupid. IPC matters. Scaling matters. Clocks alone are a very poor indicator of performance.

different architectures do different amounts of work each cycle

Shh, it's a tripshit, they're always right.

I never said any different.
But you seem to think that 1mhz != 1mhz, which is completely false.

Moreover, nVidia's GPU do more work per cycle, please see any non-biased AMD sponsored DX12 test.

>hurr this doesn't mesh with my narrative so it must be trolling
Why are red hens so oldschool pathetic?

keep backpedaling

Just report him.

He's trolling. His trip is a reference to IN Elite, who was a pro-Nvidia troll years ago.

>my admitted not very clever ruse was uncovered
sleekelite##nogodsonlymen

Have fun.

But INElite actually knew a thing or two, why pretend to be someone that doesn't know a thing?

>star ground
>PCB with strictly digital signals
are you fucking retarded? differential signals, heard of those?

>power planes under GPU
another mongrel
graphics cards have 10+ layer PCBs
there are layers with massive copper planes dedicated to power distribution alone
power planes go under GPUs on pretty much EVERY graphics card

>inelite knew a thing or two
Come on man

>093762▶
>
> >star ground
> >PCB with strictly digital signals
> are you fucking retarded? differential signal

i just saw noise and pcb in the same sentence dude chill .
digital ground is noisy regardless so it doesn't really need differential signalling.

>digital ground is noisy regardless so it doesn't really need differential signalling
get a clue
every single high speed signal is done differentially and there is a shitload of those between the GPU, RAM, PCI edge connector and the video outputs