Hi guys, I've heard that the 480 is frying motherboards

Come on Nvidia fans, get on it.

>>55340770
its on you OP where is the prof?

Facts don't matter to nvidia or its customers.

>>55340770
should i just destroy that theory once and for all ?
trips decide if i shall post what is actually going on

>>55340770
seriously dude? i'm not even an nvidia shill and from that chart alone i can see that the 960 have much smaller time above 75w

>>55341477
Had a friend just buy a 960, is he fucked?

>>55341403
They're obviously not going to be able to do that because like all electronics that follow the common safety regulations that electronics are required to follow before they can be sold in developed countries, motherboards are designed with tolerances for things like this.

>>55341500
obviously he is 960 is at 380/x level this card is 40% faster

>ha ha ha im gonna troll amd owners haha XD

>>55341499
Maybe, but the average draw over the PCIe bus is about the same meaning that if the 960 hasn't been blowing motherboards, the 480 won't be blowing boards ether.

>>55341477
>>55341499
>>55341500

Three dubs in a row. Damn.

>>55340770
I have a Gigabyte 960 and it causes my system to crash pretty regularly.

I'm certain that it's not causing damage, it's just tripping the surge protection on my motherboard, but it's still pretty annoying.

>>55341533
yeah, people that said this alone would blow motherboards are blowing up smoke, that i agree

The 960 actually gave me a shock so badly that I am now dr. doom. Thanks nvidia! You really are the way it's meant to be played!

The top graphs show 960 averaging around 55, while rx480 is on 80 watts.
Or am I not reading it right? Surely OP wouldn't debunk his sarcastic post with his own image

>>55341499
>the 960 have much smaller time above 75w
According to the second set of charts, it also spikes WAY above what the 480 does. If spiking at 150 watts is bad, what then is spiking at ~250? Also, you speaky the English very goodly.

>>55341653
i don't know, smoking 5 pack a day for 3 days every 3 month vs a pack a day for 3 years, which person will get cancer first do you think?

>>55341477
samefag destroyer of shilled nvidiots
ok lets all agree that the people that creates the standars for our pc actually fucking knows better than any idiot here or on tom..
second thing we can all agree is that CURRENTLY there is NO WAY to actually directly measure how much power you can draw from the pcie unless you have a modified pcie by your self

now the pdf that showcases the technical specs of pcie 3.0 its like a manual about women almost 600 pages no one will actually read it especially nvidiots
here is a little album (i posted already on ocn)
http://imgur.com/a/Ae4pJ
as you might all see the specs of the pcie is up to 300 watts but requires a hardware that would cost like x5 times more and a fucking liquid nitrogen loop for it not to go and orbit the moon...
now as you can see on the tech specs the guys NEVER actually give the pcie anything more than 75 watts to work with ok? can we all agree here? yes? good...
now 480 is operating at 80 watts on the pcie (since this the maximum before the mb bios thinks there is a leakage and auto shut down the pc)at any given time the card CANT pass that because no motherboard currently being sold has the capacity to stop harmonics from being fucking around instead the system will just shut down
cont

>>55341640
Stop pretending that you can somehow read an exact average from that chart because the graph spends so much time above the 75W line that it's clear the 960 averages at more than 75W over the PCIe bus.

ITT: N/v/idiots doing damage control.

>>55341699
what tom did was to calculate the draw from the 6pin along with 75 watt and have a number that physically didnt make any sense (the 300 watt)
literally he pulled more power on the 6pin than on the pcie
which by definition cant be right
i could go all technicall on your asses about how this was a totally shilled review a tom from 2012 that could make kyle blush but i wont
whoever thinks a card can pull a maximum 192 as we all saw from the french site from pcie and a 6 pin is delusional end of story this cant happen on any plane of existence that is using Ohm's law

>>55341698
Seeing how the 960 also averages over 75W over the PCIe bus a more apt comparison would be one guy smoking 2 packs a day versus one guy smoking 1 pack a day with constant spikes between 2 and 4 packs in a single day.

>>55341749
wait, are we seeing the same charts here? From first pic of the 960 charts, the dotted yellow line is at less than 60, and below that you can see, in english it says average, so maybe explain where you got the idea that 960 spends its time averaging above 75?

>>55341798
Great point. How will they defend this?

Ok the issue has been resolved.
https://www.reddit.com/r/Amd/comments/4qmlep/rx_480_powergate_problem_has_a_solution/

tl;dr the whole 75w per pcie lane is rubbish. It is only spec'd that when the motherboard boots the pcie lane is not allowed to draw more than 75w immediately. One slot can support more than 300w depending on the manufacturer.

If you didn't buy a chinese noname motherboard you are fine. Nothing will fry. Nvidiots go back2bed pls.

>>55341866
answer: they can't, and they don't really need to, as it's been caught early and unless the AIBs are idiots, this problem's been fixed already, plus motherboards are resilient. I'm just speechless over how dumb this whole comparison with 960 is, with the OP btfo-ing himself with his pic

>>55341499
time doesn't fucking matter.
it's indifferent that your boyfriend got to shove his big black dick in your ass only twice before he came. Either way, your ass got blasted real hard.

>>55341951
ehh, not really a good comparison as the time you spent in someone else's ass correlates directly to how fast you cum. From your logic, the 480 will cum on your ass sooner and harder than the 960 will

>>55341698
bad analogy.
the power surge is worse than sustained overdraw.

A power surge would be a head shot. An overdraw would be a couple of shots all on non vital points. Both can kill, but the first is more certain to do so.

or >>55341951

>>55341916
They literally disprove what you said in the comments. It shouldn't exceed 75w at all. The only mobos that will be ok are ones which receive a bios update because they can support the extra voltage. 99% of everyone else who is uses some shit tier prebuilt with a $30 mobo or someone upgrading from a 5 year old gpu will have problems with this.

You just show yourself in the foot by providing that link. Enjoy everyone using it against you.

>>55342031
lel, where did you get this fact, the school of butthurt engineer? read on spike vs surge, their difference and which one damage the most, then post back here

File: hqdefault.jpg (28KB, 480x360px) Image search: [Google]

28KB, 480x360px

All these problems like power draw and 6 pin will be fixed by AIB models right?

File: 100-MB-PB01-BR_XL_1.jpg (88KB, 1200x1200px) Image search: [Google]

88KB, 1200x1200px

>>55342072
A bios update cannot fix this. You need additional supplemental power for the pcie to prevent traces from burning up and killing you board.

Asus has molex power connectors on their flagship $500 boards, Gigabyte has sata connectors on some of their boards.

Or you need to buy an evga power boost.

>>55342072
>because they can support the extra voltage
You know nothing about electricity, your opinion is invalid.

>>55340770
"In my understanding the 75 watt isnt the maximum limit, its just the default value on startup of the motherboard. The motherboard it self sets the maximum allowed watt per slot in the "Slot Capabilities Register" which you can configure up to over 300 watt per slot. In the bits 7 to 14 "Slot Power Limit Value" you can set 250, 275, 300 and above 300 watt. This will be multiplied with bits 15 to 16 "Slot Power Limit Scale" in steps x1 ,x0.1, x0.01 and x0.001. So its up to the motherboard manufacturer and the power management on it how many watt the slot is capable of. The Specifications do define the protocol and not the hardware specs of the PCI-E slot. If a manufacturer uses better parts which can handle higher amps on the contacts and the lines, they can allow the devie in the slot a higher power consumption than 75 watt via these registers. Sadly most people doesnt even read the specifications and judge things they dont understand."

>>55342191
It literally says it in the comments of your reddit page.

>muh voltage
>muh watts

Amd fucked up. Deal with it.

>>55342273
Does that mean the 750 ti was house fire incarnate? It drew 125 watts solely from the slot.

http://www.tomshardware.com/reviews/geforce-gtx-750-ti-review,3750-20.html

>>55342219
"A standard height x16 add-in card intended for server I/O applications must limit its power dissipation to 25 W. A standard height x16 add-in card intended for graphics applications must, at initial power-up, not exceed 25 W of power dissipation, until configured as a high power device, at which time it must not exceed 75 W of power dissipation. Refer to Chapter 6 of the PCI Express Base Specification, Revision 1.1 for information on the power configuration mechanism."

>>55342117
Nope, it's over for AMD. They won't recover from this.

>>55342219
>300w
>pcie

You do realize pcie power comes from the 24 pin right? The 24 only have 2 active 12v lines, meaning it only supports 150w max.

The lanes aren't designed to draw more than 75w. 90% of graphics cards doesn't even draw more than 10w from the pcie.

960 owner here, this is actually a thing in almost every implementation of the card minus some Gigabyte models, I think.

I ended up with the Windforce model since it doesn't do that shit.

>>55342306
Shill. Reposting for visibility.

http://www.tomshardware.com/reviews/geforce-gtx-750-ti-review,3750-20.html

125 watts from pcie on the 750 ti.

>>55342300
>62 watts average 750 ti
Vs
>80 watts average 480

Yeah thanks for making yourself look like more of an idiot.

>>55342369
Yet the peaks are more extreme and very frequent. Stop shilling.

>>55342273
I wasn't even the guy who posted that reddit link, I'm just a guy who thinks you're a knuckle dragging mouth breather. The worst bit is, you still have no idea how much of an idiot you are.

>>55342345
>>55342300

>/g/
>cannot read a fucking chart and read the word 'average'

>>55342345
>750
>2016
Still no 200 watts and turning your pc into a small heater snafu.

>>55342389
>amdrone too dumb to understand that consistent overdraw is worse than occasional spikes

>>55342132
most idiotic shilling i havent seen
literally

>>55340770
>+5w continuous is worse than +150w peak
laddies and gentleman, shocker, /g/ cant into electrical stuff

File: image.png (52KB, 591x645px) Image search: [Google]

52KB, 591x645px

>>55342418
>occasional

Stop lying.

>>55342369
guise, look at the averages.
the card spikes to twice the rated specs so fast that you'll hardly get a chance to fry the circuits every 100th milisecond.

>>55342452
> POWER DRAW (1ms)

how thick can you be to not realize that overdraw measured in milli fucking seconds is a normal spike that will be compensated through the psu, while constant one like the 480 is more damaging?

>>55342393
>assblasted amdrone is assblasted

What's new.

>>55342481
>the card spikes to twice the rated specs so fast that you'll hardly get a chance to fry the circuits
Wait what? That isn't how this works. The traces on the motherboard are the least of your concern.

>>55342389
>very frequent
>chart shows a 10 second power draw
>literally spikes for milliseconds

Kys

File: image.png (86KB, 591x645px) Image search: [Google]

86KB, 591x645px

>>55342526
That's 170 seconds. Here's 10.

>all this goalpost shifting

>>55342502
that has to be nextgen shitpost
tell your handler that your instructions sheets got wrong shit in them and that you are looking like a fool online trying to convey that you know how electricity even works

>>55342511
>possible wear and tear - well within margin for overprovision
vs
>spikes - way above the rated limits

>>55342594
Exactly proving my point. Saved me from posting the graph myself.

Cheers.

>>55342610
Logic doesn't matter to shills. I wouldn't be surprised if there's some serious astroturfing going on from side green.

>>55342609
That dude is undeniably trying to say that simply because the peaks happens so fast they're of no concern.

>>55342688
>what are tolerances

Yea, constant spikes to 140 watts are way better. Motherboard determines the power limit anyway. Stop shilling.

>>55341543
Prepare for trouble...

>>55342693
Which is utter bullshit when they are that frequent.

Go to mechanical turk, you've lost shills.

>>55342418
nvidiot that thinks he has a grasp of ohm law
literally thinking consistent power draw is worse than spiking to the limit power...
consisten aka stable power draw is worse than spiking literally providing harmonics to the system is is somehow better...

we are witnessing a guy that some day he will put his finger on a plug and ask "what is electricity"

>>55342713
You do realize that the 480 spikes to as high as 150w in their testing?

Nice double standards. As expected. Amdrones defending another shit amd product. It performs bad and will kill your motherboards.

Atleast the 970 gimp card had good performance to show. This card is like 30 eur more than a aftermarket 970 and performs worse.

>>55342693
>overdrawn
>compensated

>what is amps?
>why does stuff burn?

>a 100m ohm on the 12v rails to ground for a second
>a 10ohm resistor for on the 12v rails to ground one millisecond
>which one burst first?

ayy

82W maximum average, with the 8% and 9% tolerance.

They are riding the line.

Wait... why does your videocard draw power from your motherboard? Isn't it plugged in directly to your PSU?

>>55342882
See >>55342303

Just because it can, it's still not within the regulations.

>>55342830
being so into damage control that tries to devalue the people that created pcie
http://composter.com.ua/documents/PCI_Express_Base_Specification_Revision_3.0.pdf

the fucking slot CANT GET MORE THAN 80-81WATT because the FUCKING BIOS thinks its a leakage and auto shut down the pc

now you and tom can enjoy shilling it while having zero understanding on how literally anything works..

>>55342921
It refers you to numbers that include tolerances.

>>55342927
>look at this 750 spiking to 140watt >>55342713
>gets blown the fuck out >>55342830
>b-but I was lying
>the slot can't get any more than 80w a-anyway

Kys

There's one thing that quite didn't understand yet.
Shouldn't the Motherboards themselves proihibit this kind of overdrawing? Seeing that the voltage regs for the cpu are obsessively fine tuned it doesn't make much sense that they would leave the rest of the power delivery so loosely designed. 10% or so over specs should be ok for overprovisioning I guess, but those spikes shouldn't have been preemptively choked?

Or is it something like: lol who cares about power delivery for fan headers and pci slots?

so judging from this thread we still don't know.

>>55342921
The spec is 5.5A on the 12V rail, and 3A on the 3.3V rail. No wiggle room on current.

The voltages have tolerances. 8% and 9% respectively. Totals up to ~82W.

That said, shame on AMD for using the tolerances to the limit. That's asking for trouble of low and midgrade boards, which are the fucking sector they're targeting.

>>55342829
kek

take a look at the chart the op posted and you pretty much see the same peaks, the main difference being the 480 consistently goes above 75, while the 960 didn't

as a 390x owner i never thought i'd say this but some amd fans seem to really have victim mentality. just admit the 480 is a great performing gpu with weird quirks that will be solved in the AIBs, instead of
>b-but nvidia does this too

>>55342971
one more time
http://imgur.com/a/Ae4pJ
stop trying to understand something your reality have no grasp on it
literally people like you that are scientific illiterate doesnt have any place on threads like that you are a disgrace trying to devaule everything just so your narrative fits into place

File: 1456681584195.png (223KB, 644x580px) Image search: [Google]

223KB, 644x580px

>>55343014
If 750ti/960/950 owners are not having there boards catch fire then the RX 480 isn't going to do shit.

>>55342971
Completely dependent on the motherboard and settings. He's talking about the two specific incidents of 480 over 75w out of 20+ test machines.

>>55343039
>kek
http://composter.com.ua/documents/PCI_Express_Base_Specification_Revision_3.0.pdf
learn to read for once who knows you might actually learn something NEW and stop being the equivalent of an ape that just saw a pc on the rain forest

>>55343039
The 960 consistently goes over 75w, and much higher over, it just averages a lower wattage. The floor is lower, but the ceiling is higher as far as pci-e draw goes.

>>55343065
A lot of 750Ti's have 6-pins on them anyway. Not sure about the 960/950, didn't bother with those.

I don't think people's boards are going to combust with the RX 480 installed, but it may cause issues for those that wish to use the same motherboard for 3 or more years.

We just don't know yet.

>>55343063
See >>55342303

Proceed to end yourself.

>>55343142
see>>55343126
stop evading the shitstorm that is coming to claim your gelatine that you call brain

>>55343142
It refers you to a section with tolerances included. Actual power is controlled by motherboard and/or user adjusted settings. Not the gpu. Neck yourself.

>>55343065

oem != aib you dumb fuck

>>55343250
yet another post that made everyone that posted here to lose 10 iq points

may god have mercy on our souls

>>55343250
The aibs will only be better...

>>55342506
Not him, but he's calling out the fact that you have no idea how circuitry and power draw works, which is painfully obvious from the dumbass comments you made

File: 1455258520770s.jpg (3KB, 123x125px) Image search: [Google]

3KB, 123x125px

>>55343193
See >>55342303

Still getting blown the fuck out with hard facts

>mfw dropping amdrones like like their fps in tomb raider

There are people ITT right now overriding their Skylake Vcores with 1.6v manual input value while being deeply concerned with the Rx 480 power draw.

>>55343359
i posted the technical specs from the guys that actually made the pcie standar
and yet as an idiot you still trying to justify something else..

File: AMDWIN.png (255KB, 1905x917px) Image search: [Google]

255KB, 1905x917px

>Haven't seen a single image of a burnt up board from the 480
wew lads.

File: image.jpg (47KB, 600x300px) Image search: [Google]

47KB, 600x300px

>>55343442
The shilling is reaching a head. It's actually made me more excited for the non reference 480.

>mfw nvidia wants to enter the average priced gpu range right now with their production issues

Their afraid.

>>55343528
This >>55342303 is a quote from the same spreadsheet you inbred retard. How have you not realised it yet.

>>55343660
Did you actually follow through to the referenced section?

>>55343660
you retard that states the maximum allowed on a pcie itself not the actual limit they actually have in PLACE
moron

>>55343731
>maximum allowed
>limit

Yeah so my card can hit 155mph, there's still a road speed limit.

End this once and for all.

Amd fucked up. There, we have our conclusion. Everyone can go and watch the football now.

>>55341477
Rolling

trying to find follow ups on this.
so far http://www.pcper.com/reviews/Graphics-Cards/Power-Consumption-Concerns-Radeon-RX-480

still reading the intro. don't know the conclusion yet.

>>55343831
see>>55341699
>>55341748

>>55343794
let me make say it in terms you will understand
a is a cock
b is your ass
a is having a limit of 80 watts because this is the standar
b thinks he knows better
a still doesnt draw more than this because its physically impossible to do so
c enters the game
c says the same thing
a is being honest and has a vbios problem
b is being a retard as always
eventually a is getting fed up and starts thrusting into a
a shuts up because he doesnt have any grasp of how electricity works

>>55343883
Was just about to post this. Was an interesting read.

>>55343901
Ok so we can safely say:
>>55342713
>>55342927
>>55343063
>>55343193
>>55343528
>>55343731

Is person C. At least we got that one cleared up.

>>55343940
Tldr; hes repeating reddit and doesn't know shit.

>>55343940
tl dr im actually the one that made the ocn post but oh well i guess since you have such a vast knowledge you should have already know this

>>55343883
They concluded that it's an issue. Ok, that's settled.

File: image.jpg (73KB, 640x523px) Image search: [Google]

73KB, 640x523px

>>55343940
Try again, Shitlips McShill.

>>55341699
Hey retard, I hope you realize that the 300 watts specified in the PCIe specifications is the TOTAL amount of power a card is allowed to draw. By TOTAL I mean the 75 watts from the slot (which is the maximum allowed) plus the power from any additional sources, such as PCIe power connectors.

AMD fanboys aren't just retarded, they are also illiterate.

It is right there is the specifications.
http://composter.com.ua/documents/PCI_Express_Base_Specification_Revision_3.0.pdf

Page 639 in the PDF.

>Slot Power Limit Value – In combination with the Slot Power Limit Scale value, specifies the upper limit on power supplied by the slot (see Section 6.9) or by other means to the adapter.
>specifies the upper limit on power supplied by the slot (see Section 6.9) or by other means to the adapter.
>power supplied by the slot (see Section 6.9) or by other means to the adapter.
>or by other means to the adapter.

Since AMD fanboys seem to have the same IQ count as a Coke Zero, "other means" refers to things such as PCIe power connectors from the PSU.

We had the exact same disucssion when the Radeon 6990 launched.
http://www.anandtech.com/show/4209/amds-radeon-hd-6990-the-new-single-card-king/5

> For a while now we’ve been under the impression that video card size and power consumption was ultimately capped by the PCI-Express specification. At present time the specification and its addendums specify normal (75W), 150W, 225W, and 300W PCIe card operation. In the case of 300W cards in particular this is achieved through 75W from the PCIe slot, 75W from a 6pin PCIe power connector, and 150W from an 8pin PCIe power connector. As the name implies, the PCIe specification also defines what the 6pin and 8pin power connectors are supposed to be capable of, which is where 75W and 150W come from respectively.

File: C_cWRMdkaTu3mamR5Yl1O44k16zGnXQDLDsdVYbr8WA.jpg (70KB, 600x640px) Image search: [Google]

70KB, 600x640px

>>55344198
i knew that some idiot will actually say this exact same thing thats why since i have a fair grasp of how electricity actually works i did the job for you
http://imgur.com/a/Ae4pJ

have fun
be sure to have an actually rebuttal to battle this blow

File: 1430397168041.gif (1MB, 400x368px) Image search: [Google]

1MB, 400x368px

>>55344198
>Since AMD fanboys seem to have the same IQ count as a Coke Zero

>>55344198
Pci-e 3.0 is rated for 300 watts and adjustable to that level as such in enthusiast level boards. 75 watts was the limit for pic-e 2.0.

>>55344198
>the same IQ count as a Coke Zero
I know you're trolling or whatever, but that shit is weak as fuck.

File: 1443853276247.jpg (22KB, 272x285px) Image search: [Google]

22KB, 272x285px

>mfw all these amdrones are trying to deny amd have fucked up even though pcper have actual solid evidence from testing proving it's an issue and needs to be fixed asap

>>55343920
>>55344057
That was a nice read, indeed.

I only hope that this will unleash another Fcat/frame pacing scenario that will bring more comparisons of the same metrics for all of the reviewed cards.

As most people ITT are saying, Nivida did this bfore and AMD is doing it now. Both are guilty and I wholly believe that we shouldn't ignore the issue, I've been reading on it trying to inform myself on a subject which I know jack shit of. And while my curiosity is piqued, I wish how much this offense is prevalent accross the most popular cards and preferably with their OC results as well.

File: image.jpg (61KB, 338x359px) Image search: [Google]

61KB, 338x359px

>>55344401
>mfw so many lying shills.

>>55344401
toms hardware has the best power measurement equipment around and they have it wrong
>thinking pcper an nvidia shilled pr site will be objective
>jhh will be more objective than this site

>>55344057
And also that nvidia has done the same damn thing. Yet nary a peep. Shows which side really packs the fanboys and shills.

>>55344445
Asus =/= nvidia

Tomshardware called asus out on it too but it's not nvidias fault in that case. It would be if the reference had the same issues, but it didn't. Amds REFERENCE cards are having this issue hence why it's a big deal.

If and AIB fucks up its not nvidia or amds fault. Stop acting like an idiot.

>>55344301
Can you please link me to the PDF instead? Your PDF viewer is rendering that document incorrectly (making it very hard to read) and the resolution is terrible.

It is really hard to read, but I can't see more than 75 watts over the PCIe slot being written anywhere in there.
I see lots and lots of
>75 W from the PCI Express connector
though.


>>55344445
When did Nvidia do the same? Oh right, the picture OP posted?
Sorry but it is not even remotely the same thing.
1) The 960 Strix in OP's picture averages around 50 watts from the PCIe slot. That's considerably less than the 80 watts the RX 480 draws.
2) The 960 "problem" was isolated to the Strix version. Other versions did not behave the same way.
3) The Strix card was overclocked. The RX 480 is at stock.

>>55344516
If an*

>>55344198
>>55344525
every tiem!

naturally the idiot thinks so highly of himself that he fools himself into thinking that it would a great idea to get a trip. You're spreading your stupidity, not your intelect; as all of the other narcisistic trip-fags out there.

or do you need a proof of identity to collect your shekels? Lost a lot of posts which you couldn't collect? You got jewed out by your jew overlords.

>>55344525
fuck the pdf its big
here is the sauce
http://imgur.com/a/Ae4pJ
notice the table and how they limited the pci3.0 into drawing 75 watts when a 6 pin connector is on the card..
now go back to tom
he claims the card got 200 watss from pcie
he literally said the card pulled 125 more watts than its physically possible from the very same people that actually created pcie...
and then you see pcper
"we saw the card drawing 80-84 then goes apeshit and says 6.5 amp and spikes at 7amp and that is a 27% jump"
literally we learned that a 0.5 increase is 25%you can smell the nvidia pr police there

One thing I'm getting from all this is that it appears to be questionable if even the big review sites with all the toys are capable of doing these tests properly.

Whether or not it's true, it appears that it's becoming increasingly important that the PCIe standard's testing method & equipment needs to be made public.

This isn't the first time it's happened, and it always ends up being

>This is operating outside spec.
>Says who?
>Says X.
>But Y says...
>How do we know X or Y are testing properly?
>...we don't.

If they're going to be testing for this stuff, the publisher of the standard ought to make it crystal clear how it's to be done to minimize these disputes.

>>55344728
problem is for years we believed that tom had the best machines to actually measure power draw..

today pcper showed a heavly modded pcie slot similiar to the one amd and nvidia uses to profile the cards on specific games...
i just wonder who provided them with such an expensive shit to begin with........given that it costs more than half of their yearly paycheck

>>55344401
>>55344401
>>55344401
>>55344401
>>55344401

/thread. everyone leave.

>>55344775
how can you trust tom's if they compare a 970 g1 with a reference 480

>>55344775
>pcper showed a heavly modded pcie slot

Somehow I glazed right past that image while I was reading. That is a bit damning, isn't it?

>>55344848
hey i thought the same for techpowerup till i saw they benched 1080 on
wow wotanks bf3 and fuckin crysis 3 like anyone is playing bf3 or crysis3 anymore...

>>55341500

no. it's still a solid card for 1080p.

>>55344848
All their cards are aftermarket though even amd ones. That's perfect testing. Nobody buys stock 970's or 390s but people will buy stock 480s. Testing aftermarket 970/390 is perfectly logical in my eyes and more reviewers should do it.

>>55344671
You are linking me to the same album... I want the PDF, not low-res print screen from a PDF reader that doesn't even render the document properly.

Also, did you even read what you linked? It says right there that it is limited to 75 watts from the slot.
The 300 watt limit is for the 75 watts from the slot plus power from other sources, such as PCIe power connectors.

Your argument right now is basically "well Tom says that it uses more than the spec but that can't be possible because then it would use more than the spec".
Do you understand how retarded that sounds? The problem is that it exceeds the spec.

>he literally said the card pulled 125 more watts than its physically possible from the very same people that actually created pcie...
What the fuck are you on about? How is it "physically impossible" for it to spike to 200 watts over the PCIe slot? I think you might want to look up what "physically impossible" means because you are not using it correctly.
Also, we have seen spikes upwards of 200 watts on cards before, and the RX 480 results has now been replicated by multiple sites including PC Perspective.

Stop talking about things you do not understand.

>>55345018
the pcie 3.0 draws dynamic load for of all
it limits the usage depending of the connector you have on the card............................
the more power it gets from the pins the less it gives via pcie.......
the table is showing that for a 6pin the pcie gives max 75 watts
the 480 draws 80-84 its simple
its impossible to get MUCH of it because the system will cut you off
get it? its simple you cant draw much power than its suppose too and 480 is really stretching that part on stock...now that 300 watts that tom stated on stock is just unreal stupidly unreal

>>55344525
I'm talking about the 750 ti, not the 960 strix, cuck. If it wasn't frying cheap mobos then, the 480 won't now.

http://www.tomshardware.com/reviews/geforce-gtx-750-ti-review,3750-20.html

>>55345166
Momentary spikes in power delivery ALWAYS happen
and there is nothing in place that prevents them. Nothing. At all. Every GPU does it, even CPUs do it, in many cases to a far worse degree.

Saying that it is physically impossible is an outright lie, and it makes it sound like this is your first time ever dealing with electronics.

>>55345199
but its fine when nvidia does it because you know..
its nvidia
but when amd does it its bad and should be banned because its amd

>>55345166
Can you please rephrase that in English?

And no it is not impossible for it to draw that much. Have you looked at ANY of the measurements done by the multiple sites which are all reporting the exact same thing?


>>55345199
Did you even read the article?
>64 Watts avg

Motherboards are made to handle very short spikes. Asus even responded saying that. The problem is continuous load.
Think of it as being struck by lightning. It is very possible to survive it because it is only an instant. If you however were to grab an electric fence with the same power as a lighting strike, and then hold onto it for a few seconds you would die.

>>55345275
But what if I yell SHAZAM first?

>>55345275
what multiple sites? only techpowerup and tom is the rest is just quoting them

literally 2 out of 43 reviews
>>55345224
no one said spikes cant happen
what i said is that such a number for a spike CANT happen you cant just jump 125 watts even for a ms and not having any major damage that is why we have a fucking active pfc psu the fact that he saw it and was able to continue till he decided not to overclock it shows that he was lying

>>55345340
THOSE SPIKES CAN AND DO HAPPEN
THEY DO NOT DAMAGE ANYTHING

STOP TALKING OUT OF YOUR ASS

>>55345347
prove it
show me any fucking thing that run on electricity and actually has survived over 100 w spike

>>55345166
How can one person be so retarded? Are you Greek by any chance? I've been seeing people telling you to fuck off for weeks now and you still come back with more bullshit. Neck yourself retard.

>>55345166
Do this cards have a design that pull all available power [pins+pci-e] to regs and THEN feed the board or do they lump it all together kinda agnostically?

if it's the latter it'll be tricky to correct it via software.

>>55345373
>hurrrr transient spikes in current draw are FATAL guys
Go take a community college course in circuit design, you spastic. Your fucking air conditioner will peak at couple thousand watts when it kicks on.

File: image.png (51KB, 591x645px) Image search: [Google]

51KB, 591x645px

>>55345275
Mean is not median. The 750 ti spent a lot of time over 100 watts. The lower floor influenced the average. 88 watts was the average draw for the entire card, and stock slot draw never went over 100 watts, unlike the 750 ti.

Boards and lanes have only gotten better since then.

>>55345340
>what multiple sites? only techpowerup and tom is the rest is just quoting them
Tom's Hardware.
PC Perspective
Reddit users.
Golem.de
Hardware.fr
Heise

All 6 of those are doing their own tests and coming to the same conclusion. They are not just parroting what Tom's said.

The reason why a lot of other reviews did not see this issue is because they measure power at the wall. This issue can only be seen when measuring when measuring directly from the PCIe slot.


>>55345373
The PC Perspective and Tom's Hardware articles are proof. If you want even more proof then go read the official Asus statement. It says their motherboards are designed to handle 95 watt spikes, but not continuous use at that level.

>>55345424
you surely are taking good care of these threads, arent you?

>>55345385
great, now the greek moron started his shift too.

>>55345385
Bashes on greeks. Becomes known as the greek himself.

>>55345424
toms hardware>200 watts
>pcper 6.5 amp to 7=25% increase
>reddit users-> bios problem
>golem >same as tom
>hardware.fr lvl 192 watts
>heise->same as tech
only pcper did on their own and tom the rest didnt even bother
and unless pcper explaind as to how they managed to get a profilic pcie bracket on their hand and not being from nvidia the we all must assume that is just another bias pcper article
and we already know that tom was wrong on his measurement so..

>>55345511
Literally i see this guy on countless posts shilling everyone that speaks about amd as greek

he might have lost his only gf by a greek or something i cant explain it
or his gf is working on amd and is actually getting more money than him

>>55345533
Tom's testing methodology is solid, and they have never missed a beat since they bought their current testing equipment, and did an article detailing why it was needed. For every card released since their reviews have been the gold standard for metering power usage, and they have never shown any bias. They've been totally objective about it which is why they show all their data in every review.
You are trying to discredit the source simply because you do not like their findings.
AMD's official response was that there was something wrong with the card they tested. Thats it.
Stop. Talking. Out. Of. Your. Ass.

>>55345533
Are you actually retarded?
PC Per applied a second low-pass filter which removes huge spikes.

>One interesting note on our data compared to what Tom’s Hardware presents – we are using a second order low pass filter to smooth out the data to make it more readable and more indicative of how power draw is handled by the components on the PCB. Tom’s story reported “maximum” power draw at 300 watts for the RX 480 and while that is technically accurate, those figures represent instantaneous power draw. That is interesting data in some circumstances, and may actually indicate other potential issues with excessively noisy power circuitry, but to us, it makes more sense to sample data at a high rate (10 kHz) but to filter it and present it more readable way that better meshes with the continuous power delivery capabilities of the system.


Now can you kindly shut the fuck up you fucking retard?


The Reddit user did not have BIOS problems, unless you are saying all RX 480s has BIOS problems, in which case the problem is still there.

Golem did their own tests.

Don't know what you mean by lvl 192 watts

Please explain how "we already know that tom was wrong on his measurement".


Do you not understand the difference between "spikes" and "continuous" is? It is really obvious to anyone who passed elementary school.

>>55345610
>I am Greek shitposter

Thanks for confirming. Go and rant on /v/ about how the fury x is 200% more powerful than a 980 ti.

>>>/v/

>>55345612
i couldnt care less to be honest but the results so far on ocn/beyond3d and reddit speaks for themselfs

he was wrong on this occasion no one could replicate such a power total draw on stock clocks as he said it was (300 watts)not even on beyond3d...
the only thing we got it that the card indeed has a bios problem holding the voltage up to 1.3 after the boost clock gets down resulting to more power draw...
but shit me all day if a 0.2 voltage is enough to see such a big increase on the pcie as he claims
surely it has but not that much like we need a fucking 1kw generator only to start the card..

>>55345655
Ignore him. He's some Greek guy whose been here for like 2 weeks straight and a bunch of people I know who post here have noticed him being a retarded shill every day. Just ignore him. He's probably a forever alone basement virgin.

the reviewers should just put this card in an average motherboard and keep running benchmarks for a week straight
either somethign burns after an hour or we can rest this case

Its time. Video cards need to get a 24v or 48v power bus. Drawing that much power through 12v is stupid.

>>55345678
>>55345709
being this retarted
>whoever talks about amd is greek
>whoever is greek is talking about fury x
>no one mentioned fury x
>we are talking about power measurements
>still shitposting like a true nvidiot
yesterday you came on the review thread and literally was quoting dozens of random people and calling them greek
literally every day you just come on amd threads and calls everyone a greek
being this butthurt

>>55345786
>being this greek

>>55345807
just proving my point..
you need to see a doctor
a greek doctor for your case

>>55345655
http://www.overclock.net/t/1604477/reddit-rx-480-fails-pci-e-specification/100
have fun reading

https://www.youtube.com/watch?v=8Oc7zXhzlzU
(OC3d tv)

he starts to talk about the pci power draw at around 10:00. And mentioned that AMD will come up with an official reply tomorrow (first I've heard of it).

File: 1453623189151.jpg (19KB, 251x242px) Image search: [Google]

19KB, 251x242px

>>55345786

HAHAHAHAHAHAHAHAHAHAHAHAHAHA

I wasn't even posting yesterday but I know who was posting to you and I saw what you were talking about and you confirmed it was you when you were asked. It literally is you posting in all these threads HAHAHAHAHA. You're digging your own grave here.

>pay your debts

Kek

>>55345870
sorry lend me some money for me to have a debt
otherwise no thnx im more capable of shitposting and living than you ever had..

now since im talking about amd
brb gonna buy a fury x so i can become a greek shill

>>55345904
>>55345870
Holy Christ, could both of you just shut the fuck up. I can't even tell what either of you are arguing.

>>55345923
>being this greek
PAY DEBTS

>>55340770
>The 960 also did this, spiking up even higher than the 480 to 250W.
Bullshit. Asus Strix version using GTX960 chip did this. Reference card with GTX960 chip or card by any other manufacturer didn't.

In this case, it's a fucking reference card that does that.

>>55345923
he is an idiot that comes everyday and shitposts on everyone talking about amd
literally whoever says the name amd he probably has a macro on his mouse saying "greek shill"
DAYS now

>>55345951
I'm not you're fuckbuddy, pal. Could you guys take it somewhere more private?

>>55341446
Tom's hardware tests, look it up.

>>55345923
It's some guy that me and like 8 other people have noticed keeps posting on here with amd shilling and giving stupid answers. He does it every day and we get shitty threads like these where he keeps spouting bullshit all the time and just generally shitting up decent debates.

>>55345995
Both sides are unbearable with the shilling right now to be honest.

>>55345972
Yes, and the Strix 960 was never fixed or recalled, but still didn't blow any boards because of something called tolerances. Not only that, 75W is what the PCIe standard allows you to draw at boot, just have the card BIOS modify a few bits on the motherboard's side and you can draw as much as 300W via the PCIe bus.

Sorry to break this to you, but you didn't get your anti-AMD "gotcha" that you thought you got.

>>55346006
today its greeks vs nvidia as we already have seen

apart from that
the level of shilling 2 days now only indicates that jhh will give another interview saying that they dont care about price perfomance anymore and the vfm is 20 years ahead

>>55346006
You see what I mean? >>55346038

This went from a technological discussion to this guy getting all assblasted and has devolved into shit flinging now.

>>55340770
>frying motherboards
nice fud pajeey

>>55346034
They don't care, man. They know the story. They just want those sweet, sweet rupees.

>>55346068
see what i mean of you being an idiot?
this was my first post here
>>55345533
you are literally shilling the wrong pers..err sorry greek on your situation

>>55346117
>t-this was my 30th p-post here

Ok. We know you're a deluded shill. Go and make retarded comments somewhere else.

Reddit is that way ------>

Since the card took up to 200W of power WHEN overclocked, would it work fine if I don't give a shit about overclocking and let it run stock?

pretty funny how AMD boasted about power efficiency gains, then proceeded to make an architecture that only matches nvidia's 28nm perf/watt and left off a pcie power connector for marketing reasons, leaving their card killing motherboards and itself at reference settings.

>>55346174
bahahahah literally i love calling out idiots sadly i cant see your face to laugh a bit more but oh well i will manage

>>55346194

no. the card is drawing too much power at stock. you will need to undervolt it to bring it back into spec.

>>55346196
wonder what will happen when nvidia actually manages to install more hardware on their cards to support dx12

oh wait this will never happen till volta..

>>55346231

it won't matter what nvidia does or doesn't have if AMD can't actually put out a functioning product.

novideo officially has no competition and is the only decent brand to buy a gpu from for gaymers, maybe intel or someone else can step in and make some decent dgpus.

>>55345866
it was abit odd.
ttl says that the reference cards are being sold at the upper limit of its OC potential. The clocks where they're right now (1266mhz boost) are demanding more power, and any further attempt at OC will run into the same bottleneck.

But well, the cards are running stable, as shipped, and you can squeeze even higher clocks even within the reference power constraints but they will still need some more juice.

What I found weird is how come these cards can achieve high stable clocks with what appears to be not enough power to reflect on the scores. I mean OC 101 you try to set a target, if it's unstable then maybe you increase the voltage until you hit the ceiling. You find the balance clocks x power x heat and you're set. But if understood correctly what TTL says then we have plenty of room for higher clocks somehow magically ignoring the needed voltage values to get at stable value (nonetheless they don't reflect directly in more performance until you add more power).

Software shenanigans?

>>55346034
>Sorry to break this to you, but you didn't get your anti-AMD "gotcha" that you thought you got.
You mean the thing that AMDrones tried to do on reddit? Nobody gave two shits about Strix 960 because it was only one card well into the chips lifespan. RX480 has this problem on launch with a reference card.
"B-but they did this t-too!"

Nobody sane said that it's goind to blow up motherboards.
At worst, it's going to mess up on-board sound and network card.
The problem is that Strix 960 will mess up sound every x seconds for 1ms. RX480 is going to mess it up as long as it's at load.

>>55346265
till you see this..
https://www.computerbase.de/2016-06/radeon-rx-480-test/13/#diagramm-call-of-duty-black-ops-iii-2560-1440_2

File: 1430980813601.jpg (17KB, 480x360px) Image search: [Google]

17KB, 480x360px

>>55346205
>mfw he's so mad he wants to come and see me
>mfw he's too obese to leave his basement
>mfw he's too much in debt to buy a wheelchair

Kek

KEK

>>55346219
You can undervolt it, but 75w is only the boot requirement, not including tolerances. The 750ti drew much more from the pcie slot frequently and it didn't cause problems on older budget boards. You can put 225 to 300 watt through a non shit pcie 3.0 very safely.

>>55346270

it's because cards with too few or low quality vrms can't deliver higher voltage consistently, on some cards (like gtx970s with reference pcb) you can actually get better OCs by reducing voltage with a bios mod.

>>55346270
well kyle (kyle being kyle you dont know when he is serious or trollious)said he spoke with some aibs and they said they have cards ranging from 1400 to 1600mhz
now take this shit with a grain of salt coming from kyle but if this shit is true then either amd played hide and seek or they went full retard chocking the card so much in order for the aibs to sell more

>>55341699
>we can all agree is that CURRENTLY there is NO WAY to actually directly measure how much power you can draw from the pcie unless you have a modified pcie by your self

>add an extension board
>measure current on +12V of that board
>boom you now have power measurement

>>55346310
I've read about some people suggesting a software upgrade to draw most (75%) of the power from the 6 pin connector to fix it. Would that be possible and fix the problem?

>>55346196
While I agree the single six pin was a bad decision, I don't think any motherboards will be fried, other than the shittiest of the shit. Just look at the 750 ti pcie slot draw.

File: oh the humanity.jpg (332KB, 800x803px) Image search: [Google]

332KB, 800x803px

I warn you, some users may find this imagery disturbing.
The PCIe slot exploded with such force it destroyed the both the RX 480 and the background. Only the Gigabyte Ultra Durable™ motherboard provided by our sponsor, Gigabyte, producer of the Ultra Durable™ line of motherboards resisted the explosion. What you see here is the afterimage of the explosion, still clearly visible after 3.5 hours. Truly tragic.

>>55346293
>call of duty

>>>/v/

>>55346310
>The 750ti drew much more from the pcie slot frequently and it didn't cause problems on older budget boards.

or we just didn't hear about it because shitty builds where motherboards die are usually blamed on shoddy PSUs or similar things, an OCed 750ti that drew too much from the bus could have easily killed many builds and gone unnoticed

>>55346337
have you seen an extension board with logic memory?
no stop shilling that is literally a profiling bracket one that is too expensive for them to buy and one that obviously didnt come from any retailer....

>>55346270
to try to clarify my point (some exaggerations may apply):
stock is 1266mhz @ 1.350v (firestrike p 17.5k stable)
oc is 1500mhz @ 1.350 (firestrike 17.5k stable)

but then 1500mhz @ 1.5v (firestrike 19k)

how come it got to the clock in the first place. or what did I understand incorrectly?

>>55346366
5 games on the list
chooses what he wants
true paid shiller

>>55346265
This

The reality of the situation is nvidia are the only company who make decent video cards right now. Amd has competitors but it's always at a cost of heat, power, drivers, etc.

Eg

970/390 = virtually the same but the 390 has a massive power draw and heat output compared to the 970

390x/980 = the 390x is slightly slower but it also has the most power consumption of any amd card ever and runs like the sun.

Same goes for fury x and 980 ti. There's always a downside to buying amd. Everything is cheaper for a reason.

>>55346373
>two pads where you can plug in your ampere meter
>logic memory
you are an idiot

>>55346361
750ti pcie draw averages at 62 w

the 480 averages at 82w, 100-200w with overclock.
It's literally a housefire waiting to happen, Spikes aren't the problem, AVERAGE power draw over the specification is the thing that will cause your PCI-E slot to generate enough heat that it can't get rid off and cause a housefire.

>>55346377
you are the one that slapped a fury cooler on the 480 earlier today?

>>55346359
Wait for custom cards. With proper chip, and even moreso VRM, cooling and better power connectors, I think this card could be very fucking capable.

It really shouldn't be an issue to begin with, however. Unless you have a ten year old budget bargain mobo.

>>55346366
>an OCed 750ti that drew too much from the bus

And that's why I got the one with the 6-pin.

>>55346401
cant argue with idiots sorry i rather go and do something more constructive
like everything else

>>55346405
It averages lower, it peaks much higher and more frequently. Median is not mean. This is an important distinction.

>>55346377
I mean since it can achieve the clock at that voltage while still running stable (booting, games, whatnot) how come it only starts to translate those gains in performance only after more power is supplied.

If TTL's musings are accurate, I mean.

>>55346420
Does the B150M Mortar count as a budget mobo? I've been thinking of going ATX, but found it to be too large for my needs.

>>55346421
This is why I'm getting a 480 that has an 8 pin, minimum.

Shouldn't be a problem, but I want the headroom.

>>55346451
anything above asrock is actually FINE
(i was gonna say asus too but fanboys here might kill me)

>>55346440
peaks aren't the problem, the problem is the average power draw what part of that don't you understand?

http://www.pcper.com/reviews/Graphics-Cards/Power-Consumption-Concerns-Radeon-RX-480/Overclocking-Current-Testing

Don't OC if you have a 6 pin reference card

>When we zoom in we find that the motherboard is actually providing more than 95 watts of power over the +12V line and maintains the 5 watts from the +3.3V line, proving that we are indeed getting more than 100 watts through a PCIe connection that is only rated at 75 watts. The 6-pin PCIe power cable is almost crossing that 100 watt barrier too.

>>55346444
What's the granularity on the clock monitor? If it's not very fine, it could possibly be micro-throttling from power starvation, resulting in a drop in performance at the lower voltage.

>>55346427
how about you commit suicide

>>55341500

lol, the 960 was a stupid buy when it was new. Buying it NOW? Your friend should just kill himself

>>55346392
>Everything is cheaper for a reason.

That just about sums up the 480. What a fucking disaster.

>>55346505
i rather watch you die by admitting that amd is better instead of suicide
seriously you sound like that idiot from ocn ileakstuff same pattern on shilly
same shitposting

>>55346440
>The PCI Express specification language is more specific on currents than wattage limits, calling for a maximum of 5.5A over the +12V line and 3A over the +3.3V. Our direct power measurement tools and hardware allow us to not just measure power but directly record the amperage and voltage.

> The motherboard is pulling more than 6.5A through the slot continuously during gaming and spikes over 7A a few times as well. That is a 27% delta in peak current draw from the PCI Express specification. The blue line for the 6-pin connection is just slightly lower.

The 480 is literally pulling 2 A over the specified current, there's no excuse for this.

File: 1466766951374.jpg (71KB, 264x324px) Image search: [Google]

71KB, 264x324px

>Replies: 216
>Posters: 55

>>55345373
Holy fuck you are retarded. What makes it even more embarrassing is how you've been acting like some EE-god all thread long. Please refrain from posting your uninformed trash in the future.

>>55346486
I never mentioned OC. If you don't think the 750 ti hitting 100 watts plus ten times a second on the slot alone doesn't make a difference but the 480 averaging 85 watts total draw with a six pin does, I don't know what to say. Pcie lanes have proven they can handle the 480, especially when it's within lane draw tolerance whereas the 750 ti wasn't. Ignoring that pcie 3 can handle 225-300 watts, which was less common when the 750 ti was around.

Bottom dollar, dollar to donuts, even this dog shit reference won't cause widespread problems. Extremely isolated at best.

I personally never buy reference to begin with. Bring on the OC editions.

>>55346534
But you're the one that thinks pluggin in ampere meter to your circuit requires additional LOGIC MEMORY. You obviously lack basic electronics knowledge.

>>55346591
It's already over the specifiaction at stock.

Again the 62w average with peaks isn't the same as a sustained 80+ w which is over the specification. Did you forget to mention that the current being pulled is also above the max power specifications for the slot?
You're literally pretending that the specifiactions don't exist for a reason and are strawmanning power spikes and not addressing the issue that the 480 is pulling more than the PCI-E specifiaction for wattage and current through the motherboard at stock. At overclock it's literally going to dangerous levels so you don't even have to mention it.

File: RX480_Review_Guide_DX11-pcgh.PNG.png (43KB, 606x890px) Image search: [Google]

43KB, 606x890px

>>55340770
Bullshit. But it's true that it's not much better than a 970 in some dx11 games.
I would still recommend it over a 970 bc of more ram newer architecture and probably increasing performance through driver optimization.

>>55346496
but if that were case would be very hard to miss on the reviews. It would a bigger shitstorm than this PCI-gate.

I dind't get TTL's logic that one can achieve high stable clocks on these cards at a voltage value that whilst being enough to run those higher clocks is otherwise or simulataneously too low to actually make the performance increase.

>>55346615
and this is why i cant argue with apes...

this is a profiling bracket moron hence the logic memory

>>55346652
providing a bench with highly clocked 970
saying the 480 isnt better than 970

jesus stupidity level 100

>>55346694
>I cant argue with apes
leave your parents alone then

>>55346738
you mean our parents

>>55346648
Yeah, because it's a dog shit aluminum brick cooler with no real VRM cooling and amd got obsessed with efficiency so much they screwed the pooch with the six pin. I wouldn't be surprised to see custom boards with 8+6 connectors or more, binned chips, copper pipes, and three fans OCing to levels that humble the 980 at $285.

>>55346652
Better DX12 performance and some yet to be reviewed VR optimizations are the selling points.
Supposedly some driver released yesterday already improves perf by about 5%.

>>55346757
how much more? 8x8x8x8x8x8x8 pins?

>>55346781
Two would probably be overkill, but I'd buy it.

>>55346757
Its not efficiency, its low cost.

The RX 480 would be very efficient if it were clocked in the sweet spot for GCN, which is about 850mhz, and has been for every GCN based card to date.
They just pushed clocks as high as possible to hit a performance target to roughly compete against the GTX 970 as cheaply as possible. Thats why they were bragging about the small die on twitter. They have solid margins on their $200 mid range card.

The showcase for perf/watt is low clocked Polaris 11.

File: 1288342612966.jpg (6KB, 178x200px) Image search: [Google]

6KB, 178x200px

>>55346540
no/v/dia pr shills in full force meanwhile everybody from my uni gonna or already bought a 480 about 200+~ cards just from my uni everybody I know want to buy a 480 and most just wait for AiB coolers, there are like 2-3 nvidia fanboys in my circle of friends who try to tell us the 1070 is gods gift to mankind and they get so buttblasted when we tell them to stop shilling one started screaming yesterday while telling us that 480 is trash because it can´t beat a 1070, we laughed him out of the door.

>>55346652
I love these benchmarks. The reviewer isn't stupid enough to test stock cards vs the 480 when literally everyone who bought a 970/380x/960 got an aftermarket one. However people who want the 480 are all buying stock. This is a great representation of real world performance between the cards.

Only a retard will argue against this and I bet they don't even have a reference amd/nvidia card themselves.

>>55346757
It's got nothing to do with the backplate, the card is DRAWING power from PCI-E slot, Having a more efficient cooling system isn't going to stop the card from drawing that much power from the PCI express slot. Adding extra pins also isn't going to alleviate the issue. The issue is that the card is drawing too much power through the motbo slot instead of the ATX pins. Nobody gives a shit if it goes over the ATX ratings cause most quality PSUs can take it. The problem is the current and wattage flowing THROUGH THE MOTHERBOARD.

>>55346822
>FE
>only utter retards buys reference

seems about right.

>>55346811
Oh, I totally agree there. I should say, they were aiming for efficiency, but didn't hit their performance target at that efficiency. So they upped the clocks and volts (a little overboard on the latter to save on binning, bad move imo). At that point they should have scrapped and resigned around an 8 pin, but they didn't. Big fuck up.

Still hyped for aftermarket cards if the price is right.

>>55346870
You can say the same thing about the 480 with this whole fiasco.

You don't get paid by amd, so don't shill. There's a good and bad side to every company and you just explained one, so did I.

>>55346850
I never even mentioned a backplate. Are you taking the piss? The 480 peak pulled 85 watts through the mobo. Whoop dee doo, Basil. The 750 ti peak pulled 140 through the mobo.

>>55346850
You are an electrical engineer? No? Oh thats what I thought.

>>55346955
You don't need to be an electrical engineer to know that exceeding current and wattage ratings are a bad thing.

File: x8zHnyx.jpg (173KB, 1280x720px) Image search: [Google]

173KB, 1280x720px

>>55346908
>I should say, they were aiming for efficiency, but didn't hit their performance target at that efficiency.

The die was designed that way from inception. There are no binned off CU or anything. Its 4 clusters of 9 CU, only 32 ROPs on die. At some point AMD decided power consumption didn't really matter in this segment, and 150w~ was tolerable. They kept it as small as possible, and pushed clocks to make up for the lower pixel throughput provided by 32 ROPs. Doing the most with the least die area and transistors possible.

>>55346977
>cpu free streaming and recording

Does this mean they'll finally get game stream and shadowplay? I've been asking for ages but nobody seems to have an answer.

I'm asking for a mini itx build. Kinda like an on the go pc I can move around with ease.

>>55346977
Your thinking of cut/lasered/burned, not binned.

>>55347014
Raptr or gaming evolved or whatever the fuck these days supposed have it, but I have no need for replays so I never install it.

>>55347014
They've had it forever, their VCE always had the capability. Unfortunately they paired up with that shitty Raptr service and used their software instead of offering something developed internally. Raptr failed, and renamed itself, the software is still available though.

>>55347026
When components are fused off on die, that IS binning. They are not physically removed from the die, they are just disconnected from the BEOL so they no longer draw current.

Clocking binning, voltage binning, and defect binning are all binning.

>>55347063
Binning usually precedes cutting, yes. But they are not inherently the same. Both processes cost money, if the chips are almost universally stable at a certain voltage by design (as you said) very little binning is necessary. They are going bottom up rather than top down.

Whether that is a mistake or not remains to be seen.

File: 1456886107946.png (75KB, 338x149px) Image search: [Google]

75KB, 338x149px

>>55344516
> Amds REFERENCE cards are having this issue hence why it's a big deal

1# If it didn't do anything when Nvidia cards did it then it wont do it when AMD cards do it

#2 Reference AMD designs exist for like a month or 2 and are allways phased out extremely fast

>>55344516
I was talking about the reference 750 ti, friendo, which peaked at 140 watts from the pcie many multiple times a minute, which is a bit bigger than the 95 watt max slot draw the 480 hit in isolated tests of pre production cards, or the 80 watt average total draw including the six pin. And it, the 750 ti, was powered solely by the motherboard pcie slot. No six pin or anything.

This issue doesn't matter the moment we get 8pin connector non reference cards, right?

>>55347127
custom coolers come out on the 14th

>>55347206
AMD said it was an issue with some of the cards themselves, as in they were showing aberrant behavior that shouldn't be happening.
They'll probably recall some of them and ship out ones fixing the issue.

Some people on Anandtech and a couple review sites already picked up on AMD's really loose binnings for Polaris 10. Each die gives itself an ASIC rating which vaguely indicates how well it clocks for a certain voltage. All the cards by default are running with more voltage than necessary just to ensure they could ship as many as possible.

Manually setting the voltages for all the pstates alone would bring down power consumption considerable.

>>55347206
>>55347206
Pretty much. This reference is extra poo in loo yet they didn't price appropriately overseas, and reference always sucks to begin with. Custom 480 boards with good cooling, binning, and power input will probably make the 1060 and 970/980 look silly, but 1070 is actually a good value if it ever drops to msrp.

File: AMD-Radeon-RX-480-8GB-Accelero-Mono-Plus-6-pcgh.jpg (749KB, 2560x1708px) Image search: [Google]

749KB, 2560x1708px

>>55347300
Some assumed the 8gb variant would cost 300€, so 270€ isn't bad imo. Much better price/performance than the 500€ 1070.

By the way a reviewer put a cute Accelero Mono Plus cooler on it and could push the core to 1350MHz while staying under 80 °C in games, which provided +-16% performance gain.