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

[HeadsUpHigh]

17 hours ago, Briggsy said:

Those aftermarket 8-pin, 1500mhz cards will probably pull 250 watts, just watch.

and who cares about that either ? They won't rape my mobo's electronics and that's what I care about. The power consumption means nothing in terms of cost. And it doesn't necessary mean that they will be 1500mhz.

[Trixanity]

29 minutes ago, mariushm said:

snip

It's plausible. The CU count definitely seems suspect and I've read some good arguments why there could be a "bigger" chip and that it is being held back but right now there is nothing solid, so it could just be wishful thinking on our part. The ROP count is too low as well I think. Not that a potential 40 or 44 CU chip would do anything about that.

 

Anyway, the new node is definitely at a point where it is a good idea to ship two cut down chips at first (470 and 480) until the fabs have ironed out the initial problems and until they're at a point where the full chips aren't just the few lucky ones surviving for the stockpile but they are at a point where they get good enough yields to actually start selling them. Judging from AMD's new naming convention, it seems plausible that they could launch such a chip as an RX 485 at some point (and not an RX 490, despite it being hinted at in AMD documentation) and possibly having Apple as first-priority customers as you say.

 

Of course, all speculation on my part.

 

I'm actually hoping AMD will start to get the ball rolling on the Polaris notebook chips soon (besides launching the rest of the desktop). They need to get in there - without them frying any boards of course.

[Sintezza]

Gosh this is realy getting blown out of porportion.

Let just wait for more information on this potential problem.

And if it realy turns out to be a concerning problem, then AMD will come with a statement about it.

If the power overshoot is realy going to burn out more motherboards, and we frequently get reports of it.

Then it will be a big issue for AMD, or the board partner who made the referense boards.

 

 

[i_build_nanosuits]

14nm it's got to be EPIC right?! right?!

[Briggsy]

33 minutes ago, HeadsUpHigh said:

and who cares about that either ? They won't rape my mobo's electronics and that's what I care about. The power consumption means nothing in terms of cost. And it doesn't necessary mean that they will be 1500mhz.

Was watching a livestream last night (think it was Jay and Barfacles) and Jay was commenting about how the gpu die itself may actually determine where the power is drawn from, not the components on the PCB, so It's still up in the air whether custom 480's will solve this problem or exacerbate it.

[Sintezza]

4 hours ago, Dabombinable said:

No no no, its Tek Syndicate who are AMD biased 

 

TS isnt AMD biassed in anyway either.

Not sure where you get that bullshit from.

 

[LAwLz]

1 hour ago, Prysin said:

This thread keeps going, because a certain OP has a lot of salt for a certain product. And a certain fetish for quoting specifications before he understands them and or quote a actually active and valid standard.

Most people I see quoting the specification without understanding them are AMD fanboys from what I can tell. The standard clearly stats that the maximum power over the PCIe slot is 75 watts. End of discussion.

 

1 hour ago, Prysin said:

Even among those discussions its mostly Nvidiots throwing gasoline at the fire and people coming back at them with proof that Nvidia has had problematic products too.

The problem with the "proof" is that it's an apples to oranges comparison. Spikes vs continuous use are separate issues. Besides, this thread is about the RX 480. Want to talk about Nvidia? Go ahead and make a new thread. Don't derail this one which is clearly about the RX 480.

 

 

 

45 minutes ago, Trixanity said:

@LAwLz You're just as guilty as everyone else. Don't pass the blame off. It doesn't befit you.

Guilty of what? Correcting misinformation? Guilty as charged.

Guilty of making this thread go on and on? I will put all of that blame on the people who keep on spreading misinformation.

I get that most people don't have the time or will to research a subject for 10 hours before making comments about it, but if you don't fully grasp something then maybe you shouldn't jump in head first to defend/attack a company? Maybe you should lay low until you understand the basics and then voice your opinion, instead of throwing shit and then realize that you were wrong.

 

TL;DR: Don't talk about things you don't understand.

If people followed that advice then like 4/5 of the thread would be gone.

 

50 minutes ago, Trixanity said:

You jumped in with both legs and it was glorious. It's not misinformation if you don't actually read it. Don't be mad.

I did fully read your post. The problem is that other people don't. The exact same picture you posted as a joke was actually used in a serious post just one or two pages later, to try and make the exact same point that you were jokingly making. People are illiterate dumbasses, and they will leech onto anything they see which agrees with them. Then it will be parroted over and over until people think it is the truth.

I get that you were not fully serious with your post, but I am not a fan of joking about misinformation because some moron will believe you.

 

Just to be clear. I am not calling you a moron. I do think that you were careless though, and it might have contributed to the huge amount of misinformation floating around.

Again, my only goal here is to correct misinformation.

 

1 hour ago, Trixanity said:

I still wonder why AMD wired the card to draw power equally from the bus and the aux. It was a bad decision given the power target. The RX 470 will probably do just fine (and the board should be pretty much the same). Given what I've heard of the board design otherwise, I don't think it'll be an issue if power were to be limited on the bus and all the remaining power going through the aux as the board is apparently very over-engineered (blunders aside). So what AMD should have done is limited the bus power to 50-60W and put an 8-pin on it and rebuffed all negative assertions about power consumption and said it was for overclocking but of course avoid all mentions of "overclocker's dream". That way everyone would have been happy and it would be a good pairing with WattMan as well. AMD gambled on the 6-pin and lost.

It's possible they increased stock speeds (relatively) shortly before launch to boost performance causing higher consumption than it was targeted for but for whatever reason stuck with it. I don't quite buy that but it's possible.

My guess would actually be on the latter. The card didn't perform as well as they had hoped so they increased the frequency to meet their goal. Seems more likely than a mistake at the design stage, AMD not caring or AMD being oblivious to the entire issue.

 

 

 

The forum just deleted my entire post when I pressed "show new replies". ;_;

Posts never get as good when you write them a second time.

[Sintezza]

8 minutes ago, LAwLz said:

 

 

The forum just deleted my entire post when I pressed "show new replies". ;_;

 

Most likely because of reasons?. .

people yell and scream too much in this thread.

 

And as far as Jay and Barnacules techtalk is concerned, they only trow oil on the fire.

certain parties are currently investigating it, including Asus.

And just lets wait for more information.

 

[zMeul]

22 minutes ago, Sintezza said:

Gosh this is realy getting blown out of porportion.

Let just wait for more information on this potential problem.

And if it realy turns out to be a concerning problem, then AMD will come with a statement about it.

If the power overshoot is realy going to burn out more motherboards, and we frequently get reports of it.

Then it will be a big issue for AMD, or the board partner who made the referense boards.

a couple of pages back there is a video with someone having issues running games on a cheapo AMD mobo, it shuts down

but doesn't do so when he puts a 980 on it; he even changed the PSU for 750W one

 

so yeah .. it's much confirmed at this point

[Briggsy]

4 minutes ago, zMeul said:

a couple of pages back there is a video with someone having issues running games on a cheapo AMD mobo, it shuts down

but doesn't do so when he puts a 980 on it

 

so yeah .. it's much confirmed at this point

interesting, I wonder if thats a safety being tripped in the pci-e bus itself or just a cascade failure that trips some other unrelated safety trigger?

[laminutederire]

https://www.reddit.com/r/Amd/comments/4qfwd4/rx480_fails_pcie_specification/d4tb2ev?context=3

I've seen a few guys like him which gives a little bit of hope.

That guy has literally NO issue with the card and with the non updated drivers. However, when he updated his drivers that's when he started the power draw go up.

That's pretty interesting. That may have been only luck, but it could be the sign it's a driver issue.

[zMeul]

1 minute ago, Briggsy said:

interesting, I wonder if thats a safety being tripped or just a cascade failure that trips some other unrelated safety trigger?

it's uncertain

 

from what reviewers say regarding their overbuilt test benches is that the power traces for the PEG are directly connected to the 24pin ATX power connector and have no other components on them

 

it's possible that some older mobos were built with over current / over power safety circuitry and aren't direct traces to the ATX connector

[Trixanity]

3 minutes ago, LAwLz said:

Most people I see quoting the specification without understanding them are AMD fanboys from what I can tell. The standard clearly stats that the maximum power over the PCIe slot is 75 watts. End of discussion.

 

The problem with the "proof" is that it's an apples to oranges comparison. Spikes vs continuous use are separate issues. Besides, this thread is about the RX 480. Want to talk about Nvidia? Go ahead and make a new thread. Don't derail this one which is clearly about the RX 480.

 

 

 

Guilty of what? Correcting misinformation? Guilty as charged.

Guilty of making this thread go on and on? I will put all of that blame on the people who keep on spreading misinformation.

I get that most people don't have the time or will to research a subject for 10 hours before making comments about it, but if you don't fully grasp something then maybe you shouldn't jump in head first to defend/attack a company? Maybe you should lay low until you understand the basics and then voice your opinion, instead of throwing shit and then realize that you were wrong.

 

TL;DR: Don't talk about things you don't understand.

If people followed that advice then like 4/5 of the thread would be gone.

 

I did fully read your post. The problem is that other people don't. The exact same picture you posted as a joke was actually used in a serious post just one or two pages later, to try and make the exact same point that you were jokingly making. People are illiterate dumbasses, and they will leech onto anything they see which agrees with them. Then it will be parroted over and over until people think it is the truth.

I get that you were not fully serious with your post, but I am not a fan of joking about misinformation because some moron will believe you.

 

Just to be clear. I am not calling you a moron. I do think that you were careless though, and it might have contributed to the huge amount of misinformation floating around.

Again, my only goal here is to correct misinformation.

 

My guess would actually be on the latter. The card didn't perform as well as they had hoped so they increased the frequency to meet their goal. Seems more likely than a mistake at the design stage, AMD not caring or AMD being oblivious to the entire issue.

 

 

 

The forum just deleted my entire post when I pressed "show new replies". ;_;

Posts never get as good when you write them a second time.

The whole 300W over PCIE thing should hopefully have been stomped to the ground by now. It's been going back and forth for too long now. If it really was specified to do that, we wouldn't bother with aux power. I guess that's the best reason to put forth if we want to put the final nail in the coffin without resorting to technical details.

 

As for the second paragraph: this thread is in many ways long past the expiration date. We're keeping it alive. I'm not sure if the whole spiel about not understanding the subject was aimed entirely at me, not at me at all or the thread in general including me. But let's be real here: none of us truly know what's up in this situation. We have some results but not really any idea why it does what it does and how to fix it. We're all guessing based on previous cards, the board design, the results presented and pcie specifications. AMD are the only one's who can tell why and how to fix it. Rest is just speculation. Including what you and I put forth.

 

Third paragraph: it's NOT misinformation if all facts are presented. It's the job of the reader to read it all and draw a conclusion. If they fail to do so, then yes, you could argue that they are morons. My post, as you mention, was humorous in nature based on the idea of mirroring the exaggeration and outrage in this thread on all sides. I'm poking fun at how seriously some people take these things sometimes (and I'm guilty of it myself on occasion). But on the topic of misinformation: it's not new to this forum. Frequently topics are started on misinformation. But to re-iterate: it's not misinformation if the data is there for all to see; it would have been misinformation if I took something out of context or withheld information deliberately to make my claim appear true. The only thing I could have done better was to post the entire source but it's unlikely those who would avoid reading a short post like that would read the source anyway. You can give a man a glass of water, but you can't make him drink. 

 

Fourth paragraph: as I said it's possible but seems like an oversight if it was actually the case.

 

Fifth paragraph: Indeed. It has happened to me before. The frustration alone will make you re-write the post as quickly as possible and just enough to cover what you intended to say but it's never the same.

[Kobz360]

8 minutes ago, zMeul said:

a couple of pages back there is a video with someone having issues running games on a cheapo AMD mobo, it shuts down

but doesn't do so when he puts a 980 on it; he even changed the PSU for 750W one

 

so yeah .. it's much confirmed at this point

It was a very old AM2 motherboard to be fair, as I said in that post. We're waiting for pcper I believe to test it on a cheap (but modern) board. 

[NumLock21]

They need to test out with sub $50 to $60 dollar boards to see if it will damage them slots, but I don't think it will happen right away. It might take some time for it to happen.

[Sintezza]

19 minutes ago, zMeul said:

a couple of pages back there is a video with someone having issues running games on a cheapo AMD mobo, it shuts down

but doesn't do so when he puts a 980 on it; he even changed the PSU for 750W one

 

so yeah .. it's much confirmed at this point

 

Yeah i saw that.

Would have been nice if he had a thermal laser to test how hot the traces arround the pci-e slot came before it craped out.

[zMeul]

14 minutes ago, Kobz360 said:

It was a very old AM2 motherboard to be fair, as I said in that post. We're waiting for pcper I believe to test it on a cheap (but modern) board. 

and I have a 7y old mobo I've put a GigaByte GTX970 G1 on it, I've been using it for ~1 and 1/2 years without issues

and it's a PCIe gen2 mobo

[gilang01]

23 minutes ago, zMeul said:

a couple of pages back there is a video with someone having issues running games on a cheapo AMD mobo, it shuts down

but doesn't do so when he puts a 980 on it; he even changed the PSU for 750W one

 

so yeah .. it's much confirmed at this point

that "cheapo" mobo is an AM2 motherboard. old, very old

[DevilishBooster]

@Majestic @Prysin @patrickjp93 @zMeul (There are probably others I was talking with, but you guys were the main people).

I have done more reading and research in the past 24 hours than I have in probably the last month (and that's saying something). Basically I felt like I was back in college cramming for an exam where I skipped all the classes for the entire semester. Lol!

I just want to publicly say (because I am not some unapologetic AMD fanboy, I like both Team Green and Red) I was mistaken on some comments in this thread yesterday regarding this situation. So, what I am trying to say is that...

I was wr-        I was wro-        *deep breath*        I was wwrrrrrrrrr-arrrrgh!! *sigh*

I was wrong. (Fuck, that hurts to admit... Lol)

I have been reading through the PCI-SIG specs and a bunch of articles, and I think I know where my misunderstanding on the power delivery was and I feel I now have a better understanding of how the power delivery works. It saddens me that AMD screwed up this bad because I was really hoping this would bring them back as a competitive player in the GPU market. Unfortunately, it looks like this is not something that a BIOS update can fix, so this is probably going to really backfire on AMD big time. I guess at this point we just have to hope that they can fix the issue and still bring the 480 to market so that there is a strong player in the $200-$250 range, even if it does use more than 150W.

[zMeul]

2 minutes ago, gilang01 said:

that "cheapo" mobo is an AM2 motherboard. old, very old

read my post above yours

 

and ps: AMD hasn't manufactured anything new (chipsets) for couple of years - they are still on PCIe gen2

[zMeul]

5 minutes ago, Sintezza said:

Yeah i saw that.

Would have been nice if he had a thermal laser to test how hot the traces arround the pci-e slot came before it craped out.

they're most likely in between layers

[79wjd]

3 minutes ago, zMeul said:

read my post above yours

 

and ps: AMD hasn't manufactured anything new (chipsets) for couple of years - they are still on PCIe gen2

FM2+ is PCIE 3.0

[Bleedingyamato]

4 hours ago, stconquest said:

Well, have have a first confirmed failure of a mobo due to excessive power draw from the RX 480:  Some Foxconn AM2 board.  Did you watch that vid from the Science Studio?

 

The board is fine, the system would just shut down at GPU intensive moments.

I'd forgotten to but I just did.  It's nice to see the board wasn't destroyed but it's still troubling to hear that it happened.  I'd be curious how Intel boards handle this problem.  In that video with that TwoCents guy and the other guy discussing this they'd said as a result of some cheaply made (MSI I believe it was) motherboard being used to even slightly OC a CPU it actually destroyed the motherboard for some people I believe it said.  (It was late, I was tired so I'd have to rewatch that video to double check that.)  So if I'm remembering the video right then it suggests it's possible to not simply have a shutdown but rather actual damage to a motherboard from a power related problem.

 

The Science Studio did claim that the issue with the 480 might be solved by adjusting the configuration/number of PSU connector pins so if a fix really could be that simple there may be reason to hope that non-reference 480s with sufficient PSU connector pins may not have this motherboard issue.  Assuming I'm understanding all this correctly.

 

[Kobz360]

1 minute ago, zMeul said:

and I have a 7y old mobo I've put a GigaByte GTX970 G1 on it, I've been using it for ~1 and 1/2 years without issues

Yeah I'm not denying that it's an issue man, I'm just saying that if the board was relatively modern (I'd say in the last 4-5 years), then it may not affect many users. 

 

[Briggsy]

1 minute ago, djdwosk97 said:

FM2+ is PCIE 3.0

A lot of pre-built PC's using AMD still opt for the FX6300/8300 lines, which probably haven't changed components much since AM2.