• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD RX 480 Fails PCI-E Specification

Global Foundries have been failing for quite some time, but AMD had to buy from them due to contracts.

You cant say something is bad when comparing two different architectures on two different processes.

There are numerous reasons why it uses more power, but compared to 28nm for the performance of a similar architecture it uses less power as it should.
 
I was watching AHC's stream on youtube and he was looking at the reference card.

https://www.youtube.com/watch?v=E_E2eqtm4Yw

The 12V PCI-E and 12V 6 pin are totally independent with the PCI-E connector powering the GDDR5 and the 6 pin powering the core. The over draw seems to be in memory bandwidth situations where the VRAM is working hard.

The 6 pin is actually wired like an 8 pin in terms of number of power pairs, the sense pin is used as another ground as if the 6 pin is not plugged in the core has no power and the card can't start.

He thinks there may be resistor pads that you could fill in to link the two and encourage draw from the 6 pin but this would mean the card would attempt to start without the 6 pin in and it would not know.

Certainly interesting.
 
Could this be a result of being reworked hastily from HBM? Raja said Polaris supports both. Makes sense considering they are barreling toward an inevitable HBM APU.
 
Last edited:
Could this be a result of being reworked hastily from HBM? Raja said Polaris supports both. Makes sense considering they are barreling toward an inevitable HBM APU.

If you look at the number of memory controllers in the die mockups, it has just as many as hawaii. So they can either use HBM as well as GDDR5 or just that they are also capable of using a bus that is twice the size as well.
 
I was watching AHC's stream on youtube and he was looking at the reference card.

https://www.youtube.com/watch?v=E_E2eqtm4Yw

The 12V PCI-E and 12V 6 pin are totally independent with the PCI-E connector powering the GDDR5 and the 6 pin powering the core. The over draw seems to be in memory bandwidth situations where the VRAM is working hard.
He was making a wrong guess at that point. Watch his 'Overclocking Livestream 13: RX 480 first session part 2' at 58min.

Half the vcore power phases (the VRM is 6 phases) are fed from from the PCIe connector, the other half from the 6-pin.
He thinks there may be resistor pads that you could fill in to link the two and encourage draw from the 6 pin but this would mean the card would attempt to start without the 6 pin in and it would not know.
That's a bad idea. There are good reasons the PCIe connector and 6-pin 12V are separated and you may kill something when you simply connect/short them.

The 480 has a very nice power controller, an IR3567B providing 6 phases for vcore. It is capable of adjusting the current for each phase. So, it would be possible to re-balance current draw between slot / 6-pin via the BIOS or driver. (AMD likely won't do it, since the 6-pin is already at the spec limit.)
 
The 480 has a very nice power controller, an IR3567B providing 6 phases for vcore. It is capable of adjusting the current for each phase. So, it would be possible to re-balance current draw between slot / 6-pin via the BIOS or driver. (AMD likely won't do it, since the 6-pin is already at the spec limit.)

It would be the ideal solution to the problem - bootstrap (not sure what level of reprogramming is available to that IC) the controller to rebalance the current limits on the phases - being at the limit for 6 pin is kind of meaningless as they are so over specced and there are loads of cards that draw more than the theoretical limit already - but drawing to much from the socket is a no no.

EDIT: Looks like it supports MTP, etc. so should be a breeze to reconfigure aslong as AMD haven't done anything silly.
 
Last edited:
Well, I have tried to open an RMA on my card with a different online retailer (I didn't know how else to describe to them the issue and why I want to return as the item has been mis-sold). Not sure I want to take the risk with broken PCI-E slots or alternatively neutered performance!

Rather get my money back and get a custom cooled one or wait for something else.

Do I have the right to return my item if it's not classed as faulty or mis-sold? Anyone else done this?
 
I saw that evga do a sorta pci-e power boost thing for not much money. Giving more power to the pci-e slots or something. Would something like that not help ref 480s out in any way?
 
Yeah even if its been used for a brief time, as long as all the packaging is in perfect condition and in sellable nick then they cant refuse. They might charge a restocking fee but afaik that is suppose to be illegal these days or so i have read.
 
Cheers for the info, my online request is telling me to call them now. Annoying, will have to do it tomorrow! :(

The info on the site I bought it from makes it sound like you can only return in brand new unopened condition.
 
Well, I have tried to open an RMA on my card with a different online retailer (I didn't know how else to describe to them the issue and why I want to return as the item has been mis-sold). Not sure I want to take the risk with broken PCI-E slots or alternatively neutered performance!

Rather get my money back and get a custom cooled one or wait for something else.

Do I have the right to return my item if it's not classed as faulty or mis-sold? Anyone else done this?

It is panic that has no doubt being created by parties to try and hurt AMD sales.

The facts are all modern PCI-E gaming cards can and do exceed PCI-E limits on the voltage, we have not seen any failures on motherboards from this at all for past several years. All the current NVIDIA and AMD cards can draw more than PCI-E limit and no damage has being caused.

It is nothing new, but for some reason some select individuals trying to make it sound like a bigger problem than it is and one which only effects RX 480 when in reality all cards are equally effected.

I would not worry, of course some people will still be frightened by such scare mongering and as always we pride our selves on customer service so anyone who does not want to chance it can simply return their card, but cards such as 960, 970, 380, 390 etc. can also draw more power than PCI-E specification and no one made so much noise about those cards. :)

Also AMD have already announced a fix for those who want to stay within PCI-E regulations and the driver that resolves this shall be released on July 5th, so there really is nothing to worry about. :)
 
I saw that evga do a sorta pci-e power boost thing for not much money. Giving more power to the pci-e slots or something. Would something like that not help ref 480s out in any way?

Might help to alleviate one aspect of the problem i.e. if you had a pair of cards it would mean all the power wasn't coming through the same bit of the motherboard but won't help if you are approaching the current limits more localised to the cards in terms of what the traces and pins, etc. are safe for.
 
Cheers for the info, online request is telling me to call them now. Annoying, will have to do it tomorrow! :(

The info on their site makes it sound like you can only return in brand new unopened condition.

Don't have to be sealed.

As soon as you've cancelled, take good care of all of the goods, as you have to return them in reasonable condition (but not necessarily in the same packaging) and within 14 days. If the goods are damaged the seller can take the relevant cost from your refund.
 
Well, I have tried to open an RMA on my card with a different online retailer (I didn't know how else to describe to them the issue and why I want to return as the item has been mis-sold). Not sure I want to take the risk with broken PCI-E slots or alternatively neutered performance!

Rather get my money back and get a custom cooled one or wait for something else.

Do I have the right to return my item if it's not classed as faulty or mis-sold? Anyone else done this?

Yes no quibble 14 days however you may have to front the return postage ...however as this as come to light and you where unaware of this fault 'edit' issue then the seller must arrange return postage payment of the item ...so you not out of pocket for some thing that is not your fault ....
 
Last edited:
sellers shouldnt replace the card if it is not a problem. if they are accepting returns on this issue then they are confirming the issue.
 
sellers shouldnt replace the card if it is not a problem. if they are accepting returns on this issue then they are confirming the issue.

the power draw 'issue' doesn't need to be a reason for return under the long distance trading act.

As long as he has all of the packaging and the card is in good condition, OCUK should refund it.

Once funds have cleared, there is nothing stopping the guy ordering an AIB
 
Back
Top Bottom