>8 pins fix everything
http://youtu.be/Jq47qmwcus8?t=92
>>55481988
3.5 GIGS
.
5
G
I
G
S
>>55481988
Nobody suggested that an 8-pin alone would allow higher overclocking. But the cards with an 8-pin will also come with a BIOS with a higher power target and voltage limits, as all custom cards do, which will definetly help overclocking.
((((>>55481988))))
>>55482065
No. I am not referring to overclocking. I am referring to how so many idiots screamed that 8 pins on an AIB would fix the power issues with the RX 480 and that the PCIe 75w power spec was a major issue when most decent motherboards pull more than that across the board. If you put 2 75w GPU's in you are drawing 150w from the PCIe portion of the board. It odes not matter if you draw it from a single or two slots. It's all the same to the board. Like he said. Unless you have a REALLY crappy board it's not an issue.
>>55482190
There is no power issue, esp with the updated driver. its over
>>55482221
But some are still crying about it still being 'slightly' over spec. Makes me laugh quite honestly. You can never win with trolls.
>>55482252
Yeah stay on funnyjunk retard
AYYMD HOUSEFIRES
AYYMD HOUSEFIRES EVERYWHERE
>STILL cant catch up to a 980 stock
>barely beats a 390x at those clocks
lerl still for a $200 its a good deal
>>55482221
>There is no power issue, esp with the updated driver. its over
RX480 is choking on 150Watt, it needs moe to achieve it's true potential at 1400-1500Mhz.
A guy with 1500Mhz overclock literally soldered additional wires from PSU onto the power circuit to squeeze more outta it.
>>55486072
It has though.
Guys have squeezed more than 980 outta the reference PCB with some extreme means.
I'm certain that if the Nitro 480 will be able to go at least 1400mhz it will smear the 980 in it's own Gayworks
>Rx480 sells more than 1080 and 1070 combined
Raja was right,
>>55486118
>people are poor
What's new?
>>55486123
Having an option to get a gpu or to get a gpu and save myself 500$ for cocaine and hookers i will go with the 480.
>>55486145
>not having the cash for both
lmao @ ur lyfe
>>55486149
>having more cash is bad
>>55486167
>doesn't have enough cash to not care about spending it
lmao @ ur lyfe again
>>55486200
Not him, but desu you are exactly what a poor person sounds like
>>55486226
at least i don't have to suffer owning an RX480
>>55486237
You don't have to do anything, with all your money you are afforded any choice you'd like to make :^)
>>55482000
6 PINS
P
I
N
S
>>55486118
Better Marketing DESU. Usually things that released first do better than other products.
That being said, I'm glad AMD is making a comeback.
>>55486087
You can draw up to 300W from the 6pin alone with no problem.
The Stilt released something to allow +100% power limit too
>>55486370
>You can draw up to 300W from the 6pin alone with no problem
6 pin is 75Watts
8 pin is 150watts
PCI-e Slot is another 75 watt.
You seem to be confused.
Oh i don't deny that YOU CAN, however it's not what those are made for, that will be overloading your PSU and you're not supposed to do that.
You can siply short your AC socket, doesn't mean you should.
>>55486408
Specification limitations != hardware limitations
You seriously don't know what you're talking about.
>>55486426
>Specification limitations != hardware limitations
Not the same anon but specification limitations = expected safe usage. Expecting manufacturers of plastic and metal parts to handle more than three times the current without failing or frying is like trying to defuse a bomb with a blindfold on. They might over-engineer their hardware but they aren't going to go that far.
>>55486408
6pin has two pins delivering power
8pin has three pins delivering power.
75 is two thirds of 150? Doesn't compute
>>55486426
Whoops! Hit Submit too soon. Forgot to add:
>AND if they're a cheap manufacturer making a cheap part they're not going to over-engineer it at all. They're going to make it as cheaply as possible and still fit within specs.
>>55486123
https://www.youtube.com/watch?v=Eina_dk_HgE
Maybe they just want more power?
>>55486456
>specification limitations = expected safe usage
Yeah, no. If that were the case 8pin would be nowhere near 150W rated as opposed to 6pin being 75W.
The only difference between 6pin and 8pin is 2 extra ground pins.
>They might over-engineer their hardware but they aren't going to go that far.
Considering the fact that the VRMs and power circuitry on the RX 480 make the GTX 1080 look like a cheap Chinese knockoff, you'd be surprised.
>>55486470
>need two cards to even come close to one from a competitor
Why does AMD do this to themselves? Crossfire/SLI is never advised.
>>55486426
>Specification limitations != hardware limitations
I an an engineer and i know exactly what i'm talking about.
I know everything can go 20% hotter than it is supposed to actually do, but i still wouldn't do that even knowing it can.
>>55486408
6 pin should be able to draw 150 watt fine.
The 75 watt is just super extra precaution, it won't fry from 150 watt.
The big problem is that people often buy crap PSUs which can bring some complications.
>>55486488
MultiGPU is the future. Crossfire isn't their intention, Multi-adapter in DX12 is. Also SLI is literally the same as it was 10 years ago so it's pointless for NVIDIA to even keep up the charade.
>>55486489
>I know everything can go 20% hotter than it is supposed to actually do
What the 6pin connectors are capable of handling and what the PCI-SIG spec is are completely different things. 6pin connectors are literally made to handle up to 300W without a problem.
Again, pic related, if 6pin was only designed to handle 75W then 8pin would explode carrying 150W.
>>55486408
>>55486467
holy shit nigga watch the fuckin video at least
you cunts are clueless
6 pin is 2 OR 3 12v rails (3 on the 480)
8 pin is 3 12v rails + extra ground
>>55486514
>literally made to handle up to 300W without a problem.
Ok.
Then why the fuck don't they actually do it?
Why didn't the R9 295 run on a single 6pin?
>>55486145
You got me cringe a lot. A guy with gpu does not even smell cocaine and hookers
>>55486538
The 295X2 can draw 550W-600W which is far beyond the rated 375W of the PCI-E + 2x 8pin.
Thanks for bringing that card up, it just helps to prove my point.
>>55486548
why don't other 200-300 watt cards work on a single 6 pin?
Why can't polaris draw 300 watt outta one 6 pin?
>>55486595
It doesn't need to? Why the fuck would you run something rated for 300W off something that is physically capable of a maximum 300W before problems start to occur? Not only do you completely eliminate any possible headroom for voltage spikes and overclocking, you risk fucking something up.
Are you still pretending to be some sort of engineer or have we finally exposed you for the highschooler straight from /v/ that you are.
>>55486408
A standard 6 pin connector and cable has two +12V lines physically capable of sustaining 9.5-10.5 amps depending on cable length while still being in safe margin.
10*2*12=240w, an overclocked 480 could technically run entirely off a single 6pin but the pcie slot is what powers half the VRMs.
Did you even watch the video? The guy explains everything about the 480s power delivery setup
>>55486616
>It doesn't need to? Why the fuck would you run something rated for 300W off something that is physically capable of a maximum 300W
i've seen reports that Polaris is choking on 180watts power at extreme overclock because it just cannot draw more.
Why can't it?
>>55482190
>It odes not matter if you draw it from a single or two slots
That's not true in the least bit.
The PCIE slot has 5 pins on the 12v rail, each 1.1a each.
The problem is the the 480 goes way over that limit PER SLOT, actually burning the slot that was not designed for that power.
>>55486648
>If Europeans trucks are able to drive faster than 90 km/h why don't they?
Obviously because there's a limiter in it preventing it from going above that speed.
Extra precautions to avoid any possible lawsuit
Retards thought the 8 pin would fix it, but turns out the 480 power is fundamentally flawed.
>>55486703
>fundamentally flawed
The reference card was designed to perform so much at so much power use. Nothing more and nothing less. The BIOS and power controller limits in this reference card are limited to whatever AMD decided was reasonable. The only reason there is power throttling on this, reference, card is because we have no BIOS modding tool for it yet to raise the limits.
You though are simple as fuck.
>>55486703
>it, but turns out the 480 power is fundamentally flawed.
Reference 480 is.
Nitro and everyithing else will have a proper PCB, proper power system and shit.
Here's a picture of MSI Gaming X
>>55486734
Are you saying it's ok for the 480 to use so much power over the pcie slot that it actually burns motherboards?
Not even an 8 pin saves it. Shows just how bad AMD really fucked up with the 480.
Nvidicucks advertised 4gb, get 3.5
AMD advertises 4gb, actually ships out 8gb with the 4gb bios
nvidia btfo
>>55486796
the coorporation that regulates PCI standards has said theres no problem and no change itll burn up your pcie slot
>>55486796
Read.
https://setiathome.berkeley.edu/forum_thread.php?id=74838&sort_style=&start=20
Read it and read some more.
You know nothing.
I always go for the underdog
Loving it.
https://www.youtube.com/watch?v=Q7B_zO5ICXQ
>>55486514
>6pin connectors are literally made to handle up to 300W without a problem.
>literally
[citation needed]
>>55486514
*silence*
Not surprising since literature indicates that the maximum a six pin molex can provide is 192w not 300w.
>Even though the specification allows for a delivery capability of 75 W (six-pin connector) or 150 W (eight-pin connector), the total power-handling capacity of these connectors is at least 192 and 288 W
http://www.tomshardware.co.uk/power-supply-specifications-atx-reference,review-32338-12.html
>>55486910
Lol that's not even true.
There are tons of motherboards that have already been damaged by the 480.
>>55486678
Stop buying ASSCock motherboards
>>55487253
Citations needed and no don't post that 1 shitty ASSCock motherboard from the AMD forums again.
>Hey Sanjay, did you test the power in this new card?
>Oh for Ishvara, i forgot! Raja will be mad
>Don't worry Sanjay, ship them like that, the retards will beta test it for us.
>But how will we fool them?
>Let's ask Suraj from PR
>Sell them as reference cards.
>>55481988
What's your point OP? The 480 was made to consume 130W from the motherboard, which the guy himself says isn't a problem if the motherboard isn't crappy.
>>55488253
My point is that every Nvidia shill was saying it WAS an issue when it's clearly not for the majority of people.
>>55487253
There's one mobo covered in mysterious stains and pubic hair. Which was debunked.
![02[2].jpg](https://i.imgur.com/Mk8VhlY.jpg "02[2].jpg")
