Why did AMD fuck up so bad? Doesn't make any sense how they let this thing ship.
https://www.youtube.com/watch?v=ZjAlrGzHAkI
>>55402649
It's baffling, isn't it?
I think they thought it would fly under the radar and they wanted to have the 480 reference in OEM machines which often only have a 6pin.
At any rate, the Sapphire model with 8pin is looking good, but I'm getting the 470 instead which is where the GOAT value will be at.
>hype card as having gtx 980 performance
>actually has gtx 970 performance
>has to be downgraded to gtx 960 performance so it wont burn your mobo
bravo AMD
>>55402649
They're already losing market share so they needed to put something out there to remind people that they're still around and give something for their rabid fanmob to fap over.
>>55402685
>hype card as having gtx 980 performance
They never claimed that
>>55402700
They said two of them would equal a 1080, so that means 1 should be a 980
>>55402649
This guy tested it with budget systems, it worked fine
It's an issue, but it's something they can easily fix on the reference cards probably with a slight loss in performance
Otherwise anyone with half a brain is waiting for the after market models with 8 pin connectors and better power delivery, hopefully a lot of those cards can hit 1400mhz stable
https://www.youtube.com/watch?v=LxiZqQVXuzA
>>55402714
Um no? the 1080 is not twice as fast as 980 SLI in DX11 or 12 gaming, it's only in VR vs stock 980s in games that use Nvidia's VRworks portion of Gameworks for multi-projection and Single Pass Stereo(whatever that one was called)
and we have Crossfire 480 testing that shows it just below a 1080 in most games that aren't Ashes and that are decently optimized for crossfire
>>55402677
Will the 8pin even fix the shitty power design of the board? Im hoping so but then the sapphire guy said the 480 doesnt really need 8 pin
>>55402751
We don't actually know until it comes out. Wait for reviews for sure.
>>55402751
Not 100% sure but I believe Sapphire are one of the better brands when it comes to board design/VRMs and that sort of stuff. I think MSI's "Lightning" cards are the best though.
>>55402714
No. They said 2 CAN beat a 1080, they never even mentioned the 980.
The fact that the reference overheating throttling piece of shit was between a 970 and 980 was actually nice to see. I'll pick up a gigabyte once custom coolers are out.
>>55402649
What annoys me is everything after GCN 1.0 has no RAMDAC.
I have to pay extra for performance and use Nvidia
>>55402746
>>55402954
1080 is roughly equal to to 980 SLI, therefore people deduced that a 480 would be equal to a 980.
Turns out AMD was just using Ashes of course which skewed all their results.
>>55402976
Ashes is just the best case scenario for AMD, along with the new hitman game, and anything that properly utilizes A-sync with DX12
>>55402976
Sli scales worse than crossfire and I assumed that was common knowledge...
>>55402976
mate you be trolling amd is still the KING OF FUCKING GPUs. wbhile nviviashite is still lying about key components of their cards. kek cucks.
itt: nvidia shills being paid by nividi ato shill on 4chan to shill and diminish th eimage of amd and are getting checks paid to shill
>>55403112
Yeah it's all just a big conspiracy. In fact pcper, Tom's Hardware and Science Studio are all in on the conspiracy to hurt AMD.
https://www.youtube.com/watch?v=xk24cU5ldhk
>>55402733
This has nothing to do with budget systems. It's a slot machine. I dont know what the odds of winning are, but they are not zero. Specially since mobo pcie is already ovrrspecced unlike 6 pin which is massively underspecced
>>55403361
Yeah it's already burned out pcie slots on a few motherboards.
The thing I'm worried about is that even if it works in your machine for a while, who knows how long it will last?
>>55402751
Because it basically didn't, if AMD had designed the thing properly I forget now, but supposedly they aren't even fully utilizing the 6 pin, which is safer to go ever spec on anyway.
They put the 8 pin to alleviate everyone's fears and raise the OC ceiling. I think it's a good idea.
>Shilling the NVIDIA cards badly
Reminder that AMD has open source [spoiler]BSD support.[/spoiler]
>>55402714
>>55402714
They're cutting the 1080 in half to make the 1060, which will make it less powerful than a 980. It will compete with the 480. Sli and xfire are usually not 200% performance.
All in all, sounds about right. They just cherry picked a game and cards are really good in. Dx12 gives them a big boost.
>>55403573
Yeah the OP video talks about how they could've just used more power on the 6 pin and it would've been fine.
>>55403626
I think AMD might already be out of business if not for sapphire. They should be tongue bathing their balls at this point.
>>55403395
[reliable citation needed]
I'm using that fucking card right now.
no problems.
>>55402733
I will probably improve performance actually as the boost clock will have more headroom from running cooler. Google undervolting rx 480.
>>55402733
>hopefully a lot of those cards can hit 1400mhz stable
Have any of the review samples, which tend to be choice selections, been able to get above 1300?
>>55404046
1390mhz here I think it's over 1300mhz anyways, he's hit the power limit after water cooling, so we'd need more power for more voltage, assuming it doesn't fry the card
https://www.youtube.com/watch?v=w-3fi1ovAP0
>>55404046
Nobody overclocked the review samples because they were pulling 200w at the PCI-E slot during OC
Well at least anyone with the equipment to measure such things.
>>55403039
>nviashlkjss kekcuksdct
>Babby babbling
Your tears are delicious, Pajeet. A little spicy, but not too bad.
>>55404046
>>55404062
That was the 1080, here's the 480 video, Thanks Obama
https://www.youtube.com/watch?v=2kjIHgq2zBU
>>55404065
Tons of people overclocked the 480 without issue
>>55402649
It's really not all that egrigious. They can send out a firmware/microcode update with a driver release that could do one of three things, from most to least likely.
>reprogram the card to forcibly use less power overall
>reprogram the card to draw a 55/45 6 pin/PCIe slot ratio instead of 50/50ish
>figure out some software autotune that finds the appropriate lowest voltage for stock speeds per card
>>55404116
Isn't the reason the 480 draws more than 75w from the PCIe slot is because it's already drawing the maximum 75w from the 6pin?
>>55404190
No, there are plenty of other cards that even go over the max on the 6 pin and draw only like 30w on the slot.
The 480 actually overdraws on both the slot and the 6 pin.
I have no idea why AMD decided to make it draw so much from the slot, somebody just fucked up bad.
>>55404190
That doesn't matter. You can draw more than max from the 6 pin. It's mechanically a pretty beefy connector.
>>55404190
They fucked it by splitting the vcore power between the pcie and the 6 pin, when the 6pin alone couldve done the job. Indian """""engineering"""""
>>55402649
>Why did AMD fuck up so bad
>Backup components are switched on to deal with defects on the chip
>All the backups draw power from the pci slot
>Particularly bad chips go unnoticed in power testing
>>55402649
I don't get it, the PCIe specs are quite clear on what you can draw over the bus so either they have incompetent engineers or dishonest ones.
Guys where can I buy this card?
>>55404515
Its AMD so I think everyone knows which one is the answer
>>55404444
quad quads has spoken the truth
>>55402714
Why are people like this allowed to use the internet?
>>55404515
It was tested externally.
it's because they bumped the clocks and voltage up last minute because the original stock clocks (1080mhz) were 10-20% behind the 970 across the board.
similar thing happened with the fury x, which was originally supposed to be 950 or 1000mhz but they had to oc the stock to 1150mhz in an attempt to match the 980ti's performance
>>55404526
Why would you want to buy a GTX 470? Ebay.
>>55402649
>fixed by simple driver update
>3.5GB will never be fixed
POO IN LOO
>>55404515
![Laughing-Hard-Meme-07[1].jpg](https://i.imgur.com/TRdMYpr.jpg "Laughing-Hard-Meme-07[1].jpg")
>>55405181
>AMD Driver update.
>>55402751
The 6 pin is wired like an 8 pin. You can pull like 350W from it if you want. The problem is that the card is pulling too much from the slot despite having all the power it needs available.
People are trying to reprogram the VRM controller to make it draw more power from the pin and less from the slot as we speak
>>55402649
>not waiting for custom boards
>>55402700
>>55402746
>>55402954
>>55403614
>>55404775
ITT: AyyMdrones taking their words back.
>>55402700
No they said it was 980ti killer. Overclocked it kept up with the 1080.
Now look at it. Id rather have a 390x
THE RULE IS THE SAME AS NVIDIA'S REFERENCE CARD
DO NOT BUY REFERENCE CARD
YOU FUCKTWATARD
>>55408440
The aftermarket 480s are going to be as bad, just with less cooling issues.
Overclocking....pfft check the video
https://youtu.be/ysnHdXqG7Gw
>>55408463
Didn't Sapphire got the PCI-e powerdraw fixed with their nitro 480 ? I didn't really followed what was happening in general since the 1080, I just heard there was some shit going on with every release.
>>55408523
Not sure as yet. I will wait to see what these cards are like & what fixes AMD brings out before I comment!
>>55402649
All right /g/, any mobo is kill because the 480 yet? no? Then it's FUCKING NOTHING
>>55408621
Ding ding. WRONG
https://community.amd.com/thread/202410
>>55405181
>3.5
What do you mean? I bought 970 myself, should I be worried?
>>55408662
No, I never had issues with my 970 whatsoever.
It could have been a lot worse for you, you could have bought a rx480 deep fat fryer.
>>55408686
>PS: This webm shows my OCed 970 running at 55 degrees and using over 4gigs of ram - Notice those massive 20 fps drops.
>>55408440
Wait, is this the best card of the new series?
oh look another paid nvidia shill thread on /g/
>amd says between 970-980
>paid nvidia shills overhype the card to pascal titan levels
>amd release the card where they said
>paid shills goes on rampage and starts crying 3.5tears per second
>card is doing fine using 5 watts more on pcie
>nvidia drones skyrocket it on 500 watts
meanwhile
>tom peterson just gives a flip to whoever asks for async and nvidia
https://www.youtube.com/watch?v=xtely2GDxhU&feature=youtu.be&t=2h16m23s
>nvidia forums are FULL of serious problems
https://forums.geforce.com/default/board/172/geforce-1000-series/
>defected cards all over the places
>people already crying on tech sites
>shit is about to get real
>>55408760
Can you make me a tinfoil hat just like yours?
Yep theres nothing wrong with the 480s at all.
I mean its not like AMD have released a statement regarding some issues.....oh hang on.
>>55408760
>card is doing fine using 5 watts more on pcie
https://youtu.be/ZjAlrGzHAkI
https://youtu.be/2l7ErTsvhxg
https://youtu.be/kFuYc2FHgjw
Want more?
>>55408859
if paid shills didnt overtriggered over 5 fucking watts nothing really would have gone wrong
but idiotic people always bite
>>55408884
posting pcper as an evidence its like bill gates saying that w10 doesnt collect info on you
>>55408901
>>55408890
>>55408760
This guy is frootloopy. Im guessing you use Linux as you dont want windows spying on you. All that top secret shyte you have and everything.
You opitimise basement dwelling
>>55408917
actually i have arch and 14295.rs1 version of windows
so no try again paid shill
>>55408890
So it was the shills then that make the gpu overpower and draw extra? wow I didnt know this was possible
>>55408935
Is arch a make of tinfoil? or is it a type of foil hat?
>>55402649
It is because reference 480s were orignally targeted for a 150W TDP envelope but Polaris chips didn't meet performance goals within that envelope. AMD was forced to bump clockspeeds to get their performance goal at the last minute which made the 480 into a "225W TDP card".
They didn't have the time to afford another delay for a redesign and AMD wanted to beat 1060 to $199-249 punch.
Current 480 reference cards are the unfortunate reason.
>>55408943
>overpower
literally the card was going on overvoltage on boost mode from 1.1 to 1.3 and it was failing to undervoltage back to 1.1 this was the problem
now idiots like pcper that claimed the card was stealing all the power from chernobyl by just going up 0.2 on the voltage is just stupid beyond understanding
>not having 16 or 14 pin connectors
>>55408523
Wait for the reviews to see. If Sapphire based their board on the standard AMD design, it may not actually be fixed.
>>55408981
its the best type of tin foil hats
fully programmable
fully controlable
and not a single window to upgrade on w10
>>55402685
>>55402714
and this is why you don't cherry pick benchmarks. people will take it at face value and the hype will end up fucking you in the ass.
>>55408987
So why have AMD admitted theres a problem and literally every single reviewer has come out with the same point?
Also how do videos show the actual amount the cards pull when you say this is simply total fabrication?
>>55408992
its not a desing problem its a bios problem..
>>55408641
>that guy that says it burnt the 3 pci-e of his mobo,
Either he's lying or just retarded
>>55408760
>https://forums.geforce.com/default/board/172/geforce-1000-series/
Reference 1080s are a fucking joke. They run dangerously close their power envelope at load and the stock HSF is shit-tier. Throttling is known issue with people with subpar chassis cooling.
Founder Edition = Money-grab Edition for impatience Nvidiot drones.
>>55408621
There are at least 3 cases I know of, I'm sure more and more will be coming. The scary thing about this is it could take a while to kill your mobo over time, so you could run your 480 fine for months and then bam your pcie slot is dead.
>>55408998
I need one anon
>>55409007
did you even read what i said? or just paid shills have a template for every answer?
>>55408890
It's a big deal because it goes 2 amps over the pcie slot spec.
This causes it to heat up and damage the connectors.
>>55408988
>not having 22
shiggy
>>55408901
Not sure what you're talking about, pcper is a very reputable site.
>>55409013
Then again those who have them and have them setup properly dont have issues.
Under load 60% fan 67-72 degrees.
There is no issue,no throttling.
IF you ram the gpu in and let it hit 82 with fans at 30% then yes you will throttle and deservedly so
>>55409045
>>55408988
>not having 32
>>55409050
since when pcper did actually said anything good for amd?
its a nvidia pr shill site nothing more
Guys the fix is easy, just do this to your card
>>55409082
That's hilarious considering pcper used to be Amdmb.com
>>55409094
ok today we learned that pcper is actually a very reputable site
i think i need to go and die now
>>55409094
Your Retarded anon
>>55409111
*You're
>>55409082
Funny that people used to constantly say the opposite.
You just are crying as people are being nasty about your gpu
>>55402649
OMFG!!! How retarded is everybody???
PCIe 3 can handle over 300 watts if the motherboard isn't fucking shit. If the motherboard is using cheap parts, it's the motherboards responsibility to limit the power draw to a safe limit for it, not the GPU. It's no different than a PSU that limits it's power draw, just because the hardware can accept more power doesn't mean the PSU will send the extra wattage.
Seriously, I though people who had a grasp on computers were smartish? Oooo wait, it's summer, and nvidia shills have to try to defend their monopoly.
is used 980 (palit jet stream) 1year old worth buying for 300 to 350 E?
>>55409121
>PCIe 3 can handle over 300 watts if the motherboard isn't fucking shit
No it can't.
It does 300 watts with the power supply connectors.
It only does 75 watts from the slot.
>>55409120
yes by people you mean nvidia drones aka subhumans
>>55409121
Holy shit anon, Id check before you post absolute shit again.
How very wrong you are is quite hilarious
>>55403887
Something like four reviewers had issues.
I'm not going to Google for you.
>>55409134
if you are pcper you only measure the 66 rail and call it the day
"the card has a total of 150 watt max usage"
"we instead will show you only the usage of the 12v rail"
" as you can see the card draws 80 watts from the pcie slot"
"but on the 12v rail the cards again draws 80 watts! that is a suprise we never thought we would see such a thing wow"
"now to conclude as we said the card draws 80 wats from a pcie slot that is 75 wats
and also draws 80 wats from one rail that is 66 wats"
"that means the card is drawing a wooping +120% more"
"nvidia the way its meant to measure gpus"
>>55409183
They measured both, straight from the article:
"The highest power draw I measured with the RX 480 at stock settings showed 80-85 watts of power draw at over 7A on the +12V line and 4.5-5.0 watts of power draw on the 3.3V line. "
>>55409187
>>55408621
> it doesn't kill the entire motherboard, just the PCIe slot, so it's not a problem
Yeah alright then. There are reported cases of it burning out the PCIe slot. Fucking Google it you goddamn knuckle-dragging, window-licking pants-on-head retarded faggot.
>>55409156
Out of four reviewers actually able to test the issue. 100%
>>55409202
go to page two
see the table in the middle of the page
tell me if they are actually have the combined measurement
tl dr they dont they just one the 12v rail while showing the card having 80 watts that from their own words was from the combined power draw of the pcie (75 watts)
>>55409134
>>55409146
I suggest you guys actually do research, and look up specs. PCIe x1 is only able to put out 75w, the later generations can put out more and more if the motherboard allows it, and doesn't use shitty connections.
>>55409224
who? the idiot that used 4 of them for bitcoin mining on an old as shit mobo and fried his 24 pin connector?
or the idiot on amd forum that actually claimed to had played witcher 3 for 7 hours while his steam account was saying 0.3 ?
>>55409234
Dude I quoted the fucking measurement in my post, do you want me to do it again?
Here you go:
"The highest power draw I measured with the RX 480 at stock settings showed 80-85 watts of power draw at over 7A on the +12V line and 4.5-5.0 watts of power draw on the 3.3V line. "
>>55409238
No, you are the one that needs to do your research.
https://en.wikipedia.org/wiki/PCI_Express#Power
The absolute maximum the slot provides is 75W. The power supply connectors is the only way to get additional wattage.
>>55402649
Fuck my currency
>The Reference RX 480 4GB is just $10-15 cheaper than a non Reference 970
>The Reference RX 480 8GB is just $15-20 cheaper than a non-reference 980
That means the Non-reference RX 480 will be roughly the same as the 980 Non reference in price and the 4gb Non Reference is a more expensive than the 970
What to do /g/? I already sold my R9 380, I need a gpu
>>55409277
paid shill go where i told ya and see the table dont avoid the inevitable they fucking intentionally removed the second rail end of story
>>55409303
no you moron the absolute maximum of the pcie is 75 watts ON BOOT
not in general
http://composter.com.ua/documents/PCI_Express_Base_Specification_Revision_3.0.pdf
dont fucking post stupid things
>>55409307
No they didn't, re-read my post.
>>55409324
Nope, you might want to re-read that.
75W is the absolute maximum from the slot, which is why no other GPU goes over it...except the 480.
>>55409343
>>55409324
¿¿¿So if the max power it can deliver is 75w, why are all GPUs able to pull more than 75w from the slot???
>>55409325
i guess one of the tasks of paid shills is to avoid the truth kinda like jhh does
but given that the internet never forgets
>>55409343
i just gave you the formal pcie spec from the people that actually created it and you are saying THEY ARE WRONG? being so in damage control that just shills on things we know for ages?
>>55409343
> 75w is the absolute max
> except when it's not
Then 75w isn't the absolute max, is it?
>>55409386
kek reminds me of those typical nvidia graphs
>>55409377
Because the motherboard lets it.
That is the problem, the pcie slot is built for 75W, but the 480 is drawing 90W through it, which damages the motherboard.
>>55409387
75W is the official pcie spec.
The problem happens when a video card goes over that spec like the 480
>>55409377
Depending on how you look at it bad design form the motherboard manufacturer (not using a current limited supply to each slot with a hard 75w limit) as the spec would suggest, or good motherboard design, allowing a buffer above the 75w limit set out in the spec to accommodate bad GPU design.
>>55409397
yesterday it was 80watts
today its 90watts
give it a week it will reach stargate atlantis zeron point module levels
>>55409386
Good graph, thanks for showing just how bad the power problem is on the 480
>>55409415
PCper says total is around 90W at stock.
Tom's Hardware actually got it up to over 200W overclocked which is really bad.
>>55409397
Which was originally my fucking point, if a motherboard uses non shitty connections, it can output as much power as it wants, because PCIe-3 can! If they can't handle the extra out put due to cheating out on connections, it's the motherboard's responsibility to cap it at 75w...
Since the 480 turn out to be a flop is the MSI Radeon R9 390 the best deal for under $300?
>>55409415
> the RX 480 pulls over thirty-thousand watts through the PCIe slot ALONE, while running slower than a 420m and hotter than an F-22 Raptor's afterburners
>>55409435
The slot is built to handle 75W, not anymore.
That's why when the 480 goes over the limit, it damages the motherboard.
>>55409435
All motherboards should cap at 75w. That's the spec limit, regardless of the quality of components used.
>>55409417
thanks that you agree tho
at least now we know that you are saying pcie slot is 66 watts
paid shills being this into damage control
>>55409432
they literally said 80.5 see >>55409386
yes tom actually got it on 200 watts are you insane? first of all it was 300 and on torture mode
you are a paid shill but not as good as the other one although you both trade blows on stupidity
>>55409463
66 watts is just one rail, it can hit over 100W on the full slot
>>55409410
>(not using a current limited supply to each slot with a hard 75w limit)
The protection circuitry is last resort and the last line of defens, and hard limit at 75W will make many graphics cards not work because there are peaks caused by switching transistors on the leading edge of a cycle and those peaks can go above 75W.
>>55409386
Resorting to having to say everyones a paid shill as you cant give them a proper reply which explains things.
You dont get out very much do you.
>>55402649
The new PS4 is supposed to have a paralitic version of this card. How they're gonna make it?
480 is a meme card. RX 470 is the real product here $150 half price on the 1060, $100 less than 8gb 480. Solid upgrade over 280x/380x. Smaller chip with less ram so it will comfortably OC with a 6pin. CF = 980 ti.
>>55409502
Well he does use arch linux, so not going out a lot is almost a given
>>55409474
If graphics card manufactures can't work to the specs then they shouldn't be in the business, controlling transients isn't a big deal electrical engineers have been doing it for a very long time on much more complex loads than a GPU.
>>55409505
Take the normal one and get it very very drunk?
>>55409463
https://youtu.be/kFuYc2FHgjw
Shows Toms article (164w draw)
>>55409502
im still asking how the table is saying the pcie spec IS 66 WATTS
while
a) he clearly said some lines above that the card was drawing 80 watts from both rails
b) he clearly wasnt showing this on the table thus created a false image
>t seems clear at this point that the new AMD Radeon RX 480 does in fact draw more power through both the motherboard PCI Express connection
http://www.pcper.com/reviews/Graphics-Cards/Power-Consumption-Concerns-Radeon-RX-480/Overclocking-Current-Testing
and then it goes and completly fucks up our iq by showing the pcie having only 66 watts while he has the card with the combined power of 80
simple as that unless someone is to believe that the card on stock was getting 97 watts and on oc 119 watts
>>55409524
are you seriously saying the card was getting 164 watts from the pcie ?
>>55409556
Probably overclocked
Tom's actually got it to 200W, then stopped testing because they didn't want to damage their motherboard.
>>55409569
are you retarded? do you hear your self? 200 watts from the pcie slot!
not even pcper said such a stupid thing
the card is rated at 150watts at nominal gaming usage and now you say the card was getting more power than this only on the pcie?
some people are just ...
>>55409600
It's right in their review:
"We skipped long-term overclocking and overvolting tests, since the Radeon RX 480’s power consumption through the PCIe slot jumped to an average of 100W, peaking at 200W. We just didn’t want to do that to our test platform."
http://www.tomshardware.com/reviews/amd-radeon-rx-480-polaris-10,4616-9.html
>>55409622
Did they say that the 200w was a sustained 200, or simply a several-millisecond-long spike from a cycle switch?
>>55409640
200W spike, 100W average
>>55409622
you do know the card has a 6 pin connector right?
you do know that it gives 75 watts only from that right?
if we go with your numbers the its 175 watts on stock
275 on oc
and 375 on torture
do you even know how stupid this sounds? you are suggesting that this card is getting more power than the msi 390
>>55409556
164 watts total
>>55409600
200 watts total
are you retarded or something?
>>55409648
Then that's less of a problem. The spike isn't the issue. It's the sustained amount that leads to an increase in heat, which in turn causes thermal failure, that's the problem. PCIe slots routinely pass more than 100w through in millisecond-long bursts with no damage.
My only issue with this card is lack of Windows 8 drivers
>>55409673
That's typical for AMD. I had to wait 6 months for Vista drivers with my 280x.
>>55409666
100W sustained is a huge amount, because the slot is only rated for 75W.
>>55409695
Yeah, I know. But I'm saying the 200w spikes aren't the massive problem that people are making it out to be. People are running around like retards flapping their hands saying "IT'S PULLING 200w SUSTAINED THROUGH THE PCIe SLOT!" which simply isn't true.
100w on a 75w-rated slot isn't good, no, but it's not the 200w on a 75w-rated slot that some people seem to think it is.
>have a very minor driver issue with my 290x
>decide to go on amd support forums for a little help since it was beginning to annoy me
>faggots pour in screaming at me how I was a paid nvidia shill just for asking for help fixing my drivers
What the fuck is wrong with you people? You're all over /g/ too, a fucking infestation. You don't look good to the customers for the company you're shilling for free for, you know.
>>55409717
Nobody said it was 200W sustained.
The problem is at stock clocks it hits 90 watts average, which will burn out the pcie connection
>>55409781
According to AMDummys, everything is just one big conspiracy against AMD. Even legitimate customers like you become the enemy if they say anything bad.
>>55409665
are you actually retarded toms said
164 watts total
and
300 watts on torture
stop being an idiot
>>55409695
who the fuck said the card was getting 100 watts sustained from the pcie?
>>55402649
just wait a little bit you dipshits for the non-reference models.
>>55409916
This is /g/, we can't fucking wait for anything here, and we will wildly speculate and throw shit right up to the MINUTE before specs and benchmarks are released, and even past that point.
hey guys can i make an afterburner for my pci slot?
i think i have the thermal capacity
>>55409121
That is not at all how power regulation works. Brush up on your ohms law. A power supply turns an ac supply of 110-220v into, typically, a +/-12vdc, 5.5, and 3.3 rail. Nothing more.
The card itself will have methods for turning these voltages into something it can use, incorporating methods anywhere from the humble voltage regulator to the more advanced pwm controllers for better efficiency.
Without said regulators, the power supply would just attempt to supply its voltage to anything with a low enough resistance for the current to flow through. Assuming a low enough resistance, it could even end up supplying enough current to damage itself.
If all the regulatory circuitry was within the power supply itself, all designs would have to cater to whatever supply characteristics the particular manufacturer set. Which is completely impractical.
There is a small chance that they can alter the timing of anything utilizing pwm with a bios update to help shift the power burden onto the 6pin, but without detailed knowledge of how their power regulation circuit works (and im sure its quite complex, considering its a graphics car) i couldnt tell you how feasible that is. An aftermarket company like sapphire or powercolor, however, could probably redesign the entire power curcuitry if they so desired. Otherwise we wouldnt end up with balls to the wall cards like te devil 13 and superclocked variants that manage to pull 1.5x the refence card's initial power evelope.
On the subject of the excessive power draw on the pci slot, damage would depend on a number of factors in board design. The quality of the caps, resistors, regulatory components, thickness of the tracings, etc all come in to play. More than likely any decent motherboard can handle current spikes, considering they last for microseconds at a time, and the mean power draw is more or less around the target.
>>55409809
It was one of the worst experiences I've ever had with support for any device
>USER ERROR
>JUST WERKS FOR ME
>USER ERROR
>IT'S YOUR HARDWARE
then when I just said, ok, well can you help me figure out what _I'm_ doing wrong then, they started calling me a shill
>>55410048
>things that definitely happened
>>55409915
Tom's and pcper both got that when overclocking
>>55410093
Gonna say he's a paid shill next? You faggots are unbelievable, that's not exactly a farfetched story, you can probably find several examples of the same thing right now just by browsing their forums.
>>55410112
please provide actual numbers of them having 100 watts on the pcie
not some random vague statement NUMBERS
>>55410139
Just read their articles, it's all there.
>>55410142
yeah no one said anything that the card had 100 watts on the fucking pcie
it was 80 from both of them
>>55410131
The guy you responded to must be some counter-troll actually trying to get people to hate AMD and he's doing a good job
>>55410158
I already quoted it above
>>55410179
yes i asked for actual numbers not a vague statement
but its ok its hard to come by with numbers that they dont exist
>>55410131
Ive had trouble with my 390 too. No one on the forums called me a shill , no one called me a nvidia cuck , they helped me figure out that my graphics card was pulling too much power from the pci slot , so I gotta change my entire fucking motherboard.
So it actually is your hardware. AMD cards work just fine when you have optimal hardware. I originally had a 5850 in there and it just recently started fucking up. The only paid shills out there are nvidia cucks shilling their expensive hardware.
Sorry but if I had the money to pay 300$ just to avoid a bunch of weird quirks coming from my GPU I probably wouldnt have those quirks anyway having fixed all the other hardware.
>>55410202
The numbers are from the article
>>55408641
You guys do realize that the ones who fried their mobo have outdated mobos right? And all of them are AMD AM3+ AMD BTFOOOOO
>>55410229
nope they arent
>>55410266
Yes they are
>use a outdated architecture from 2012
>based on AyyMD
>slow as duck amd cuckX muh 8 corez
>try to oc both rx 480 and amd shitX
>complain about rx 489 frying pci slot
It's like you guys are begging to be cucked
>>55410280
not a single article from the 4 people that actually have done the tests
never EVER said such a thing
>>55410344
It's right in the article, I already quoted it above
>>55410224
I didn't have problems with the card, or my hardware it was a minor issue that I fixed thanks to the help I got on overclockers forum(flashed to a different bios)
My issue was just with the absolute faggots I dealt with on the AMD forums. Maybe your experience was different from mine, but you can't deny that they exist, they're all over /g/ too.
>>55410383
did you ate a copy paste burger?
>>55410400
Let me guess, English isn't your first language
>>55410421
no sorry if by english you mean bullshit that no one actually said then yes
>>55410431
Let me guess, you aren't very smart
>>55410488
depends compared to you? obviously yes
>>55410501
>obviously yes
you just fucked yourself up m8
>>55410515
lmao
>>55410515
you miss the whole point but ok as i said
>obviously yes
>>55410397
Yeah well these people are all over the internet I guess , wherever there are fanbases of opposing camps there are people arguing that the other camp sucks dick and will passionately tell you about it every time you bring up your opinions , whatever they might be.
Was your problem random sudden computer shutdowns ? When something too graphic intensive happens on screen like a particle heavy explosion , my fucking computer shuts down. Apparently you gotta buy the EVGA power boost to provide stable power to pcie slots but I think ill just buy a gimmick that will end up frying my hardware somehow.
>muh 1080
>muh 1070
>uber dx12
>nvidia can do async
http://www.overclock.net/t/1604750/dsog-report-total-war-warhammer-runs-27-slower-in-dx12-on-nvidia-s-hardware/10
epic
>>55410591
>mfw Nvidia owner and I want DX12 to succeed.
>>55410591
what the fucking point bring nvidia dx11 performance down to match amd dx12 .
What can I get on par with 480 for price and performance?
I'd get one but they seem to be sold out everywhere.
I got myself sapphire RX480. Fps at very least doubled compared to my old 7870. My PSU is OCZ 550w. Motherboard is 4 years old cheap Asrock B75. No problems whatsoever.
>>55402649
OH SHIT
They had to overclock it at the last second to make it rival the 970. I really want AMD to succeed, but they just keep fucking up.
>>55411575
The card in particular
>>55411575
>07.07.16
Whaat?
>>55411593
I'm assuming Fresh Windows 10 install and misstyped the date
This basically sums up the RX 480 fiasco.
They lied about the TDP and gave it a 6-pin instead of an 8-pin. The reference card is one of the biggest disasters in a while. Sure AIB card might be great, but AMD should get sued over the reference design.
>>55409077
winning
>>55411667
Who cares
My card arrives in 2 days and I bet it will have no problems at all
>>55411722
Okay good luck. I actually don't want AMD to fail, I just hate that they keep making stupid decisions.
All they had to do was rate the card at 170W TDP and give it an 8-pin. Instead they lied and misrepresented the card to trick low-end buyers. It's a good card, but not an efficient one. $200 is a great price though.
>>55411779
>Instead they lied and misrepresented the card to trick low-end buyers
Apparently they did that so people with 6-pin power supplies can still use the card, because it's "for the masses" and they wanted as much compatibility as possible
>>55411779
it's only okay when NVIDIA blatantly lie about TDP right
>>55409121
>PCIe 3 can handle over 300 watts if the motherboard isn't fucking shit.
At least read the fucking specs instead of a reddit thread you minge. the PCIe 3 spec is slot (75w) + 6pin (75w) + 8pin (150w).
>>55411722
Isnt like you would even say if you did have problems.
>>55409781
>What the fuck is wrong with you people?
Battered Person Syndrome.
>>55411810
>>55411802
They could have lied about TDP and even just left the 6-pin. The true fuckup and "scandal" is that it draws over 66w sustained from the PCI-E 12v rail. That's not acceptable in any situation.
>>55411921
>That's not acceptable in any situation.
t. anon with absolutely zero electrical knowledge prior to this whole situation
>>55411667
they lied about not having 75 watts and instead have 80 wats!
WE ARE DOOMED
AMD IS THE CAUSE OF GLOBAL WARMING
AMD CAUSED CHERNOBYL
AMD THIS
AMD THAT
meanwhile on the best company's forum
https://forums.geforce.com/default/board/172/geforce-1000-series/
fried dp ports
black screens
cards wont boot with certain resolutions
random crashes
nvidia good
>>55411955
I mean that's not acceptable for a video card, no matter what. Whether it's a budget card or a top-end GPU, it shouldn't be drawing significantly over spec from the PCI-E slot.
>>55411993
you literally don't know what the fuck you're talking about. why are you still posting and parroting shit you don't understand at all?
>>55411894
What?
>>55411955
If a new line of cars is bursting into flames at random, you don't need to be able to fix the engine to know that's not right.
>>55411921
anon dont know that the rails are dynamicly balancing loads due to the ocp
anon constantly brings the 66 watts of the pcie and intentionally forgets the 3.3v second rail because it doesnt suit his needs
meanwhile >>55411667
actually have it right
>>55412005
>take photo of car that burned down
>THERE IS SOMETHING WRONG WITH THIS CAR MODEL I WANT A REFUND AND RECALL ON ALL MODELS THEY ARE ALL DEFECTIVE
>>55411999
Okay, you want to explain why exceeding the spec throughout gameplay is totally fine? Because the fact is the spec limit is 66w from the 12v.
The 380 drew less than that, so it's not like I'm trying to tarnish AMD's image. It's just this card that's bad.
>>55409009
He said one slot is dead and there are burn marks on the rails. Completely consistent with concerns.
>>55412051
lets be clear on something paid drone
pcie since it was 1.0 it was never only a 12v rail never EVER
it was always one 12v and one 3.3v
now that you learned someting new lets try to debunk your shitty missleading posts
you never go and measure only one rail and call it a day its wrong at so many levels that hurts to even read it
toms/adoredtv/random french site they all concluded that it was running on 80 watts 5 watts above the 75 that is the current standar of pcie
now take your pcper shillness out of here or start talking like a true hardware enthusiast
bring only one rail and claim that it was drawing only on that 80 watts its stupid beyond comprehension
also 1.8 amp above it spec will only fry a north korean motherboard or an asrock that has 3 power phases less than even a no name chinese one..
>>55412161
>also 1.8 amp above it spec will only fry a north korean motherboard or an asrock that has 3 power phases less than even a no name chinese one..
Oh okay. So AMD is out of spec, but it's okay because motherboard manufacturers account for shitty hardware like AMD's. Got it.
Face it, they should be running more power through the 6-pin and less through the PCI-e. Like literally every other card in the same TDP range.
>>55408417
>No they said it was 980ti killer
Where. All you idiots who believed the hype keep saying AMD made these claims but have set to show a single piece of evidence to support your claims like the idiots saying the 480 was on par with the 980/980ti
>>55412227
yes you found it
because surely only amd shitty products causes harmonics and spikes on a system
>being this stupid that hurts to even answer
>>55409507
You'll be stuck at 1080p though and no amount of dx12 or driver updates will change that. Sounds like you need a job famalam
>>55408621
>can sustain 90w of draw when overclocked
>PCI-E slots rated for 75w of draw
>Motherboard manufacturers say they can handle maybe 95w of draw according to pcper
>AMD hitting 100w sustained draw from PCI-E when OCed in benches
If AMD doesn't have a way to fix this so the excess power can come through a 6+2 pin then the reference models are finished in regards to overclocking. You need to undervolt before overclocking in a bunch of cases.
>>55412435
90 watts became 100 watts on just one day
by the end of the week paid shills will increase it to fukushimal level of energy
The 4gb Version cost as much as the 390 in my country (316 Dollar)
The 8gb version cost as much as the 390x in my country (360 Dollar)
Which one is the better buy?
You niggas need to grow up and ignore nvidia shitposters. Im haveing a blast with my card and dont give a fuck if AMD sales will get hurt by nvidiots spreading misinformation.
>>55412515
Also, non-reference version.s I'm thinking picking up the XFX 390x or the MSI 390. Not sure which one is the better buy
The denial of these ati drones is getting really really sad.
>>55412515
>The 8gb version cost as much as the 390x in my country (360 Dollar)
then you should buy the 390X senpai
>>55412435
Undervolting on majority of RX480 will save you around 30W and give several other benefits. I see no reason not to do it on the reference card.
>>55412515
The cheapest 970 (3.5Gb) cost the same as RX 480 (8Gb) in my country.
>>55412000
i think he means it's gonna be hard to post anything when your house burns down.
>the cheapest reference rx480 8gb in croatia
I wish I was a cuck so I can enjoy getting fucked over and over agian by this fucking socialist/communist government and their insane tax system
>>55412850
cant u just import it from krautland?
anyways i got a 4gb one for 180 something £ and its alright desu
>>55412877
>pay german taxes
yeah, no thx. I still like my country more than germany.
>>55412850
The non reference models will be even more expensive
Never buying amd ever again. They put my life on the line to get performance their product should not attain.
![e37d0a3f[1].png](https://i.imgur.com/OIl55oU.png "e37d0a3f[1].png")
>>55412896
what, you mean the ones that still don't cause it to be more expensive?
>>55412896
on a second thought, how much does it cost in Germany?
>>55404526
In a trash can
>>55412937
>33 eur less
I dont think its worht bothering importing from other country, specially since I have to pay transport costs.
but I'm waiting for reference cards anyways. Will see what the prices of those are.
>>55412919
>tfw my mobo exploded after a 5mhz oc on my rx480 and my cat is now blind
t-thanks raja
>>55412985
yeah, that's a marketing trick family. most OC cards draw more than they are supposed to from PCI.
