AMD once again violating power specifications? (AMD RX-480)

[Majestic]

8 minutes ago, TheRandomness said:

The 295x2 disagrees. So does the 480 apparently. Plus, it's only 15W in total (ish)

150W* > PCIe bus + PCIe power cables

*could* I can draw a total of 400ish W with my 290 when it's overclocked to hell with an 8 pin and a 6 pin. 400~ > 375. 15W difference, no harm done. In fact, I'm considering pushing the wattage higher to see how far I can go. No damage done to anything so far, the only thing stopping me being thermals (This'll be my overclock for gaming btw)

That they need a 650-750W PSU c:

 

BTW: I know this is kindof a 'my card does it so it's fine' post, but my point is that is the spec is broken by like 15W, is it going to set my house on fire?

Couple of things wrong with this. 

A: you're confusing choosing to violate specifications by modifying the card with a product that violates them out-of-the-box

B: it's specifically about the PCI-E, not about whether the power supply can handle it. 

 

And yeah, it is one of those posts. And tom's hardware didn't say it will set fire, specifically said it won't. Please read what is written, besides the title.

8 minutes ago, ace_cheaply said:

What they say makes no sense.

The draw through the pci-e reached a peak of 200W, yet the card never draws more than 168, so even assuming none of the 168W comes from the 6 pin connector, that is still less than their peaks.  

The card draws 300W peak, of which 155W goes over the mainboard and the rest over the PCI-E PEG connector. Again, please read what is written.

These are fractions of a second though. The average is 82W. It might not cause components to fail, but the bursts could cause audiopopping or other perhipheral issues.

 

The >100W during overclocking might though...

[ace_cheaply]

3 minutes ago, Majestic said:

Couple of things wrong with this. 

A: you're confusing choosing to violate specifications by modifying the card with a product that violates them out-of-the-box

B: it's specifically about the PCI-E, not about whether the power supply can handle it. 

 

And yeah, it is one of those posts. And tom's hardware didn't say it will set fire, specifically said it won't. Please read what is written, besides the title.

The card draws 300W peak, of which 155W goes over the mainboard and the rest over the PCI-E PEG connector. Again, please read what is written.

Like i said, it makes no sense. I'll wait for it to be repeated.  I pulled quotes from the article, obviously I read more than the title. It says 200W peak through the pci-e. LOL. Also, note the edit. 

[Dionyz]

11 minutes ago, Majestic said:

I can tell from your rabid responses, "doe effe normaal man". Also, by vague I meant it's nothing past an aknowledgement of the allaged issue. They're not confirming or denying it.

Here. Click play

 

[Majestic]

1 minute ago, ace_cheaply said:

Like i said, it makes no sense. I'll wait for it to be repeated.  I pulled quotes from the article, obviously I read more than the title.  LOL. Also, note the edit. 

What exactly doesn't make sense?

[TheRandomness]

2 minutes ago, Majestic said:

Couple of things wrong with this. 

A: you're confusing choosing to violate specifications by modifying the card with a product that violates them out-of-the-box

B: it's specifically about the PCI-E, not about whether the power supply can handle it. 

 

And yeah, it is one of those posts. And tom's hardware didn't say it will set fire, specifically said it won't. Please read what is written, besides the title.

The card draws 300W peak, of which 155W goes over the mainboard and the rest over the PCI-E PEG connector. Again, please read what is written.

I give up, I never win arguments. No matter what points I make.

Probably not, but it's the principal.  You simply don't do this.  Specifications are there for a reason.  What if your USB ports put out 6 volts instead of 5?  There's also something to be said for integrity in advertising.  What if your car consumed 20% more gas than it claimed?  Plus, OP made a great point about many people who will probably run these probably have very cheap motherboards which won't handle this as well as some might.

Can we have a good point rating back please? Thumbs up doesn't really cover it.

Oh no! My graphics card draw 25ish more watts than advertised! My electricity bills will go up so much!

[Majestic]

8 minutes ago, Ryan_Vickers said:

 What if your car consumed 20% more gas than it claimed? 

They usually do

[Trixanity]

Only a single reviewer (Tom's) has experienced this particular problem although that could be because they have better equipment and testing methodologies. A peak of 155W over the bus when the card has a measured consumption of some 160W ish seems odd. That could indicate that the card fails to draw power over the 6-pin PCIe connector and it would be preferable for the card to take any excess from the the 6-pin instead of the bus.

 

Maybe a BIOS update can fix the power management unless it's a hardware issue and it also depends on whether it's a defect on Tom's card or a defect in the overall design meaning it affects all cards. If Raja is to be trusted, then it was never noticed in internal testing. So make of that what you will.

[Majestic]

5 minutes ago, TheRandomness said:

Oh no! My graphics card draw 25ish more watts than advertised! My electricity bills will go up so much!

Nobody is talking about electricity bills. It's a very specificly defined issue, this strawman is utterly pointless.

 

@Trixanity I expect AMD to know how to fix it, but I'm pretty sure it will hurt performance as it has to throttle more. But ofcourse, that will be after all the reviews are out. 3,5/4 stars AMD *golfclap*

[Crazy Ginger]

9 minutes ago, TheRandomness said:

The 295x2 disagrees. So does the 480 apparently. Plus, it's only 15W in total (ish)

150W* > PCIe bus + PCIe power cables

*could* I can draw a total of 400ish W with my 290 when it's overclocked to hell with an 8 pin and a 6 pin. 400~ > 375. 15W difference, no harm done. In fact, I'm considering pushing the wattage higher to see how far I can go. No damage done to anything so far, the only thing stopping me being thermals (This'll be my overclock for gaming btw)

That they need a 650-750W PSU c:

 

BTW: I know this is kindof a 'my card does it so it's fine' post, but my point is that is the spec is broken by like 15W, is it going to set my house on fire?

Probably not, but it could damage your motherboard and processor, and other pics devices like for instance, a soundcard, network card, or even an expensive pcie NVME drive.  Basically, if it does damage your mobo, you better hope that only the mobo dies, as a damaged motherboard can easily damage anything connected to it.  If a short circuit occurs, theres a chance that it will send too many volts to your processor.

[Sakkura]

27 minutes ago, SteveGrabowski0 said:

I can't believe how much power this card uses. When I saw the performance numbers this morning I was impressed, but that was back when I was thinking this was a card that typically uses 110W in gaming loads, not freaking 165W. So it's a slightly more power hungry GTX 970 released 21 months later. Except the 970 was 28 nm, so AMD hasn't gotten their performance per watt under control at all. Where does this leave Vega? 300W to match a 1080?

 

What a disgrace to see them use a 6-pin power connector out of spec to push the narrative of this being some great power efficient card. And then the bait and switch by advertising it as a $199 card and only sending $239 review samples with higher clocked memory to the reviewers.

The GTX 970 uses 170-175W, so 165W on the 480 is a little less, not more.

[vanished]

4 minutes ago, TheRandomness said:

Oh no! My graphics card draw 25ish more watts than advertised! My electricity bills will go up so much!

It's not the electricity cost I'd be concerned about.  It's the ability of boards to handle this abuse, and the principal that (assuming this all pans out to be true) they lied about the power draw and broke specifications.

[ace_cheaply]

4 minutes ago, Majestic said:

What exactly doesn't make sense?

13 minutes ago, ace_cheaply said:

 

The draw through the pci-e reached a peak of 200W, yet the card never draws more than 168, so even assuming none of the 168W comes from the 6 pin connector, that is still less than their peaks. We are to believe that without overclocking it reaches a max of 300W?  200W peak through the pci-e, and 100 from the 6 pin?  I don't believe it. If it's true, it should be repeatable and we'll know soon enough. 

 

I might not be getting a 480 after all. 

[raphidy]

12 minutes ago, TheRandomness said:

The 295x2 disagrees. So does the 480 apparently. Plus, it's only 15W in total (ish)

Great, because it has been done, doing it again is perfectly fine? No thanks, I'll skip those product/version of goods.

[Sakkura]

2 minutes ago, Ryan_Vickers said:

It's not the electricity cost I'd be concerned about.  It's the ability of boards to handle this abuse, and the principal that (assuming this all pans out to be true) they lied about the power draw and broke specifications.

They didn't lie any more than Nvidia did with the GTX 970 and 980 TDP.

[Majestic]

4 minutes ago, ace_cheaply said:

 

But it peaks 300W..... I think you misread the results...

 

 

middle is peak, top is min, bottom is average. MAINBOARD 12V : 155W. Not sure how much more simple I can make this.

[Crazy Ginger]

2 minutes ago, Sakkura said:

The GTX 970 uses 170-175W, so 165W on the 480 is a little less, not more.

It draws that power through the power connectors, NOT THE PCIE SLOT.  there's a reason why you have to plug in those wires coming straight from the PSU to your gpu

[vanished]

1 minute ago, Sakkura said:

They didn't lie any more than Nvidia did with the GTX 970 and 980 TDP.

That doesn't make this any better, really.  Best thing you can say for that is if you've been using one of those all this time, and want to encourage the continuation of this kind of behaviour, then you shouldn't worry about running a 480 - your board will handle it.

[Majestic]

3 minutes ago, Sakkura said:

They didn't lie any more than Nvidia did with the GTX 970 and 980 TDP.

Those came with 2x 6-pin connectors. TDP is up for interpretation, the cards however did not violate power specifications.

[TheRandomness]

1 minute ago, raphidy said:

Great, because it has been done, doing it again is perfectly fine? No thanks, I'll skip those product/version of goods.

I think I'm the only one here who has the mentality of 'who cares?'.

2 minutes ago, Ryan_Vickers said:

It's not the electricity cost I'd be concerned about.  It's the ability of boards to handle this abuse, and the principal that (assuming this all pans out to be true) they lied about the power draw and broke specifications.

Oh no, a company lied. That's a first.

3 minutes ago, Crazy Ginger said:

Probably not, but it could damage your motherboard and processor, and other pics devices like for instance, a soundcard, network card, or even an expensive pcie NVME drive. Basically, if it does damage your mobo, you better hope that only the mobo dies, as a damaged motherboard can easily damage anything connected to it.  If a short circuit occurs, theres a chance that it will send too many volts to your processor.

If you have a PCIe SSD and an RX 480, you put your money in the wrong places. Also, with the SMCs on the motherboard they have their own specs, meaning that it's possible for certain components to run even with the increase power draw and whatnot. Kinda like how VRMs can go to 125C whilst the GPU itself has a separate thermal limit.

 

I should stop arguing, I'm never going to win.

[ace_cheaply]

5 minutes ago, Majestic said:

But it peaks 300W..... I think you misread the results...

 

 

middle is peak, top is min, bottom is average. MAINBOARD 12V : 155W. Not sure how much more simple I can make this.

Quote, directly from the article:

 

"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. "  200w peak through the pci-e slot.  Not sure how much more simple i can make that. 

[Sakkura]

1 minute ago, Crazy Ginger said:

It draws that power through the power connectors, NOT THE PCIE SLOT.  there's a reason why you have to plug in those wires coming straight from the PSU to your gpu

That's irrelevant to what I was saying. He claimed the RX 480 uses more power than the GTX 970, when the fact is that it uses less. The PCIe slot issue is an entirely separate discussion.

1 minute ago, Ryan_Vickers said:

That doesn't make this any better, really.  Best thing you can say for that is if you've been using one of those all this time, and want to encourage the continuation of this kind of behaviour, then you shouldn't worry about running a 480 - your board will handle it.

It certainly means you shouldn't treat AMD any worse (or better) than you treated Nvidia for doing the same thing.

 

(just referring to the overall power consumption, not the distribution)

[Majestic]

1 minute ago, TheRandomness said:

I should stop arguing, I'm never going to win.

Not with 95% of your statements being logical fallacies, no.

[vanished]

2 minutes ago, TheRandomness said:

 

Oh no, a company lied. That's a first.

 

I know you're being sarcastic but it certainly isn't the first time.  We can all help try to make it the last time though by sending a message about them doing this.  Maybe companies lie because every time they do, the product sells anyway and people move on and just expect it to happen again?

[Sakkura]

2 minutes ago, Majestic said:

Those came with 2x 6-pin connectors. TDP is up for interpretation, the cards however did not violate power specifications.

And we were talking about the power consumption vs. TDP, not the issue of exceeding the specified maximum power draw from the PCIe slot.

[vanished]

2 minutes ago, Sakkura said:

It certainly means you shouldn't treat AMD any worse (or better) than you treated Nvidia for doing the same thing.

Yes very true