RX 480 is a failboard because of AMD marketing. Delusional in thinking that "appearing" more efficient by using fewer power cables and drawing from the motherboard. Guess what, you just killed your product right out the gate:
People already reporting burned PCI-E slots already: https://community.amd.com/thread/202410
Who are they hiring to make these decisions?
The whole mobo burning thing is an exaggeration/nvidia shilling you idiot. Though I agree that 480 is a massive disappointment.
>Who are they hiring to make these decisions?
I think we all know the answer to that question.
>>55379262
Didn't this pajeet in OP picture work for apple?? Seems like Pajeet don't like to test products before they release them...
>>55379553
More likely it's Pajeet's boss who thinks testing is a waste of time and money
>even PCI-SIG saying it's daijabou
>nvidiots still sperging out
its obvious the 480 was meant for poorfags
it meets the minimum VR spec just like AMD said
It doesnt have 980 performance because that was what Nvidia shills were spamming so Nvidia could pull the "dissappointed by AMD? 1060 has 980 performance" campaign
>>55379262
>burned PCIe slots
Ahh yes, a burned slot from going 5-10w over spec.
That's what we call 'a fucked motherboard'
Same thing would've happened with a GTX 960, a fermi-era card or even my own old HD 6870 which could run purely off the board if you forgot to plug in the 6-pin (aw yee).
Currently at 1315mhz on 1.10v (50mv down) with a full 150mv taken off the RAM at 1.0v and stock 2000mhz.
130w, feels good man
>>55379674
pajeets boss is a pajeet fa.m in case of Radeon
>GTX 480 was an failrue
>RX 480 is an failure
Would that mean that RX 840 will be successful card?
>>55380656
The RX 480 is only a failure in nvidiots delusional dreams.
It's already sold more than the GTX 1070 and 1080 combined in less than a week
>>55379262
>buying reference cards
Sure, AMD could have used a 6-pin, but AIB cards are already going to use a 8-pin. It's not ruined. At least it has good OSS drivers.
>>55379262
What do the red down arrows men in this picture?
>>55381196
mean*
>>55381196
That nvidia is shit.
>>55380835
too bad AMD is gonna lose a lot once they have to do recalls and driver downgrades
I thought the 480 was targeting a lower price bracket, which its currently dominating in. What was the hype about?
>>55381215
Except the only recalls will be the overclocked XFX one and the driver 'downgrades' will be a drop of both the core and memory voltage that, as far as us 480 owners can tell, not a single card on the market would have a problem with. It'll run exactly the same factory clocks, and still have unlocked overclocking for those of us who have managed a significant undervolt and overclock.
I've got 1315mhz @ 1.10V, this is actually a pretty shit one in terms of silicon lottery too.
>>55381232
The fact that overclocked they're beating 980s in DX12 games.
>>55381196
10/10
>>55379262
>democratize
Exactly the political wording I would expect from AMD
The RX 480 is literally 200 dollars. I love nvidia, but come on now you idiots. Everyone but this board is praising the card for being a great budget card. You didnt expect it to be a 1070 or 1080 killer now did you? Oh wait I guess you did.
>>55381215
>>55379262
>NVIDIOTS DON'T HAVE ANY MORE SHIT TO VOMIT ONLINE SO THEY CLING ONTO THIS TO SHILL LOL!
>>55381302
Did people actually believe that though? Just look at the price, if they were capable of doing that they wouldnt sell it for half as much.
>>55381354
AMD are pajeets, not jews
>Visit Amazon UK
>See "paid review" tier comments like "10/10 best card I've ever owned"
>One of them is literally a copypaste from anandtech.
>£219.99 for ref oven
>970 non ref costs £199.99
...Should I just wait for 1060? 760 atm.
>>55379262
>Read trough the whole thread
>Literally just ONE guy with a dead PCIe slot
>Using an Asrock board which is known for failing PCIe connectors
Then again I guess this is just what you should expect from fanboys. Making a mountain out of a molehill is not only common, but standard practice.
>>55381581
>>Literally just ONE guy with a dead PCIe slot
Second post states:
>I have this exact same issue, totally killed my 990FX from MSI.
How badly can you fail at basic reading comprehension?
>>55379553
This pajeet was in charge of the internals, mainly CPU or GPU
>>55382788
>>55379674
https://www.amd.com/en-us/who-we-are/corporate-information/leadership/raja-koduri
AHAHAHAHAHAH
MACFAGS BTFO
Problem is solved.
It isn't the GPU die drawing the power in errant fashion, it is the memory. AMD's software as well has trouble with undervolting and seeing an actual reduction in power consumption so long as memory settings are set to auto. If you undervolt the GPU then the card will give more power to memory by default.
Don't use the auto settings in Wattman. Manually undervolt, and adjust memory settings as well. Problem solved. AMD will issue a driver fix and statement next week.
>>55381196
In a world full of bullshit it is nice to see that fud and marketing do not always work.
>>55381196
That their position has fallen since last week or so
>>55381196
Deprecated designs.
>>55381286
FinFET production doesn't operate like planar transistors - like when Intel processors moved to FinFET, voltage increase to clock increase are much closer to linear than before. The silicon lottery spread is much lower now. I'd expect 1320Mhz at 1.1v to be a decent result, and many average cards would be unable to undervolt the stock turbo down to 1.0v
Very high end, highly binned AIBs might, maybe, be able to reach 1500Mhz at 1.2-1.25 but you're looking at beyond 225w power draw (8 pin + PCIe slot technical limit) with the memory overclocked. I guess the rumors of 8+6pin AIBs could be true.
Now...I'd be willing to spend ~$270 for a monster RX480 but only because I buy a GPU every 3-4 years, and that AMD's cards tend to outshine Nvidia's over time.
>7Tflop FP32 for around $250
>>55381634
>new account
>one post by this user
How badly do YOU fail at general comprehension?
>>55384193
>new account
Move a lot of goalposts? How much time did you spend digging to come up with this new goalpost?
>>55381196
that there will be lots of buyers remorse soon
>>55385692
**and fried MB's.
>>55381415
>1060
Save up an extra £150 and get a 1070.
>>55381196
How do we even know the difference between the 3rd party releases? They all look the same and don't list any differences.
>>55379262
Reference 480 is just a disappointment and clearly rushed to the market to beat the 1060.
It is wired with "out of spec" 6-pin PCIe connectors (it has three full 12V pins instead of two) which the board itself doesn't even use properly.
For some silly reason AMD made the board draw power from both PCIe and only two of the PCIe pins on the "out-of-spec" six-connector.
Somebody fucked up badly at AMD.
>>55381634
The problem comes from pre-PCI 3.0 boards.
Ironic that this issue plagues mostly AMD platforms. You have go to Sandy Bridge and older to get PCIe 2.0 slots.
>>55379718
I dunno about you, but my Fermi had 2x6 pins, and it's just the GTX 460. My GTX 980 has 2x8 pins.
>>55381196
