The 480: power consumption, PCI-E powerdraw

[redzo]

A video from pcper about this:
https://www.youtube.com/watch?v=ZjAlrGzHAkI

 

[Erenhardt]

Obviously the cooler and board were designed for something closer to 110-120W than 150-160W. It would be interesting to see how the card does if it was downclocked to, say, 1150 MHz, with concomitant voltage reduction. I bet that perf/watt would go way up.

Nope. Cooler is ok for the card. And board is over the top:
https://www.youtube.com/watch?v=qG2e-v94L4M

Those VRMs could supply two polaris 10 chips without a problem, maybe even 3. Granted you would need more PCIe connectors

The real question is, can you change the voltage controller behavior with software. Can you throttle PCIe phases down with new bios?

 

[William Gaatjes]

You make it very hard to believe you are electrical engineer.

First, good luck burning connector due to excessive current. First think that would go pop would be traces or power regulators.

Short spikes is what kills connectors/pins.

The reason for burned connectors is most of the time human error, or sometimes damaged connector. To burn the pins like shown in the photos above, you don't have to draw any excessive currents beyond specifications.

Bad contact between two pins and even light load will make it burn. Add current spikes to that, and you have a recipe for a disaster.

I've seen a 10m cable 1.5mm2 @220V go pop at the connector with 1<kW load on them. Bad contact, sparks, smoke, tears.

On the other hand I've used 40meters long 1.5mm2 @220V with welder plugged in 2kW+. The welder had problems because of Vdrop, cable got pretty warm and the 16Amp fuse went off from time to time. But in this example the cable had no connectors other than your regular socket/plug.

I am sorry, but if a connector would not burn because of excessive (overspecification) current for a prolonged period of time, there would be no reason to specify a maximum current. The manufacturer keeps in mind that the contact resistance can vary a bit. Of course spikes can cause issues too, but you would need very very high current spikes constantly. It is all related to contact resistance between the wire and the terminal and from terminal to terminal and terminal to another wire or pcb. Everything has resistance(terminals themselves too). Lot of current is power dissipation.

P=I*I*R.

 

[Bacon1]

I can't believe AMD is in such bad state they didn't understand this themselves. I am sorry but it makes no damn sense.

What are you talking about?

They said that it is giving too much power. They made a whole new power management software and it wasn't setting the settings properly. They are fixing the issue in a driver / software update which will tell it to use less power.

Lots of people are undervolting their cards and getting great results:

https://www.reddit.com/r/Amd/comments/4qupw4/super_psa_all_rx480_owners_please_attempt_to/

 

[William Gaatjes]

What are you talking about?

They said that it is giving too much power. They made a whole new power management software and it wasn't setting the settings properly. They are fixing the issue in a driver / software update which will tell it to use less power.

Lots of people are undervolting their cards and getting great results:

https://www.reddit.com/r/Amd/comments/4qupw4/super_psa_all_rx480_owners_please_attempt_to/

That is a relief.

 

[torlen11cc]

GTX 960 STRIX has the same problem:

 

[Flapdrol1337]

Any confirmed reports of this thing killing motherboards or causing damage of any kind? I'm curious to know how many older cards may draw more than 75W from the PCIe slot. I do agree that an extra 10W is most likely harmless on most motherboards.

One guy said his pci-e slot is dead, but he was playing for 7 hours straight and also overclocked the card. (so he probably increased the powerlimit, which can make power use go over 200W).

 

[SPBHM]

GTX 960 STRIX has the same problem:

seems to have been proven false

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

 

[zinfamous]

What are you talking about?

They said that it is giving too much power. They made a whole new power management software and it wasn't setting the settings properly. They are fixing the issue in a driver / software update which will tell it to use less power.

Lots of people are undervolting their cards and getting great results:

https://www.reddit.com/r/Amd/comments/4qupw4/super_psa_all_rx480_owners_please_attempt_to/

This is cool to see. Hopefully some of the big review sites take the time to go back through their benches after fiddling with the volting and clocking their 480s >1300mhz like these users are doing. That would be really informative.

Though I imagine they will probably wait for more info/official release from AMD...and of course it's a holiday weekend. :\

 

[guskline]

A video from pcper about this:
https://www.youtube.com/watch?v=ZjAlrGzHAkI

redzo thank you for the posted video.

For all of the posters claiming that the Strix 960 does the same, per the video it clearly does not exhibit the same problem.

Until an AMD official fix, I would undervolt a RX480 a bit.

Gosh, I have used AMD since the 386-40 days but to cry that this is made up by Nvidia fan boys is really putting your head in the sand.

The good news is that AMD has acknowledged the problem and is actively working on the problem.:thumbsup:

 

[Erenhardt]

I am sorry, but if a connector would not burn because of excessive (overspecification) current for a prolonged period of time, there would be no reason to specify a maximum current. The manufacturer keeps in mind that the contact resistance can vary a bit. Of course spikes can cause issues too, but you would need very very high current spikes constantly. It is all related to contact resistance between the wire and the terminal and from terminal to terminal and terminal to another wire or pcb. Everything has resistance(terminals themselves too). Lot of current is power dissipation.

P=I*I*R.

While it is true that the connector can take finite number of amps, you design your connector in such a way that they can handle more current than the line and power circuit is designed to support.

But, if your contact is wrong (due to human error, or pin damage/ whatever) and arcing can occur, the supported current drops significantly and depending on the contact patch, you may experience fireworks without any meaningful load.

 

[William Gaatjes]

While it is true that the connector can take finite number of amps, you design your connector in such a way that they can handle more current than the line and power circuit is designed to support.

But, if your contact is wrong (due to human error, or pin damage/ whatever) and arcing can occur, the supported current drops significantly and depending on the contact patch, you may experience fireworks without any meaningful load.

I agree, but arcing means a loose and intermittent electrical connection. And then it is not a matter of current spikes or high constant current.
Because the repeating contact causes arcing and causes the plating(if present) to damage and the core material of the pin itself. This can increase contact resistance making it worse even when currents flow within the spec of a normal terminal. Also, with intermittent electrical contact, the terminals can on the point of contact weld together. The opposite effect.

 

[DarkKnightDude]

Maybe the voltages are higher than they should have been? Read several reports people are able to undervolt their cards and got great results.

 

[William Gaatjes]

A video from pcper about this:
https://www.youtube.com/watch?v=ZjAlrGzHAkI

I like how he explains about the ringing. He keeps it simple but accurate.

 

[rainy]

Maybe the voltages are higher than they should have been?

It is a well known fact, that AMD has been for many years setting conservative (high/too high) vcore for theirs CPU/GPU.

 

[MrTeal]

While it is true that the connector can take finite number of amps, you design your connector in such a way that they can handle more current than the line and power circuit is designed to support.

But, if your contact is wrong (due to human error, or pin damage/ whatever) and arcing can occur, the supported current drops significantly and depending on the contact patch, you may experience fireworks without any meaningful load.

Well, in a lot of systems you would design the connector to be overbuilt especially if it's point to point, but that's not always the case in some of these systems. Some PSUs have per-connector current limiting, but not all do and you can melt a 6 pin Minifit Jr even if it was in good condition by running it over spec long term when there's 100+A on tap behind it. Similar things can happen with the PCIe slot; obviously the PSU can supply more power than a connector can handle, and the ATX connector can also handle more current than a PCIe slot with how they distributed the pins. It has to be, because you could be pulling PCIe spec from a couple slots simultaneously.

 

[VeryCharBroiled]

As someone else stated, if your card is drawing more than 75W from the mobo RMA it.

how is joe consumer supposed to check that?

heck i even have a multimeter but would not be able to do that without a special adapter or some such. even then...

 

[Erenhardt]

I agree, but arcing means a loose and intermittent electrical connection. And then it is not a matter of current spikes or high constant current.
.

Not necessarily. With very small contact patch on the pin caused by a damage or some debris, and high amperage the arcing can occur if the distance is small enough.

It happens quite often that the wall socket has a loose connection to the wire. You can plug in any device and have it working OK, but there will be arcing on that loose connection despite some contact. Most of the time it will result in burned wire and no contact.

This effect can be observed with electric welder set at too high amps with damaged electrode coating (insulation). Even direct contact at the end of the electrode to the welded metal will not prevent the arcing from other nearby spots where the electrode is exposed.

and the ATX connector can also handle more current than a PCIe slot with how they distributed the pins. It has to be, because you could be pulling PCIe spec from a couple slots simultaneously.

Yep. 24pin atx power connector need to supply multiple x16 pcie slots with power. Drawing 100 watts from the motherboard will not cause a 24pin atx power connector to burn Maybe if you did 4way crossifre with rx480

 

[William Gaatjes]

Also, if the contact is not good at the input circuit because of intermittend contact in a connector, and you have smps buck converter delivering for example 100W to a load, you get spectacular arcing on the pins. Because the smps will try to maintain the proper voltage at the output. The input voltage lowers, so the input current increases.
smps uses reactive components like inductors and capacitors. Worst case, the smps will blow up.

edit:
Best case is that the soft start circuitry of the smps takes over. But then the gpu will be unpowered for a short time and your pc will not like that. A crash.

 

[William Gaatjes]

Not necessarily. With very small contact patch on the pin caused by a damage or some debris, and high amperage the arcing can occur if the distance is small enough.

It happens quite often that the wall socket has a loose connection to the wire. You can plug in any device and have it working OK, but there will be arcing on that loose connection despite some contact. Most of the time it will result in burned wire and no contact.

This effect can be observed with electric welder set at too high amps with damaged electrode coating (insulation). Even direct contact at the end of the electrode to the welded metal will not prevent the arcing from other nearby spots where the electrode is exposed.

When it is loose, it is not making proper contact. So you have some resistance there that is higher than usual. Add in temperature differences causing materials to expand and contract, add in vibration from the environment and the point of contact heats up enough to become a plasma. the plasma moves away and the arcing starts.

With wall sockets, you have AC current. That makes arcing worse if i am not mistaking.

 

[Sweepr]

As you know, we continuously tune our GPUs in order to maximize their performance within their given power envelopes and the speed of the memory interface, which in this case is an unprecedented 8Gbps for GDDR5. Recently, we identified select scenarios where the tuning of some RX 480 boards was not optimal. Fortunately, we can adjust the GPU's tuning via software in order to resolve this issue. We are already testing a driver that implements a fix, and we will provide an update to the community on our progress on Tuesday (July 5, 2016).

www.anandtech.com/show/10465/amd-releases-statement-on-radeon-rx-480-power-consumption

This begs the question, what kind of software fix is this? Will it hurt Radeon RX 480's performance even further (or not)?

 

[Erenhardt]

www.anandtech.com/show/10465/amd-releases-statement-on-radeon-rx-480-power-consumption

This begs the question, what kind of software fix is this? Will it hurt Radeon RX 480's performance even further (or not)?

For the user feedback it improves performance.

 

[PeckingOrder]

https://bitcointalk.org/index.php?topic=1433925.msg15438155#msg15438155

it's ok, guys

just a minor setback. it's a 200$ card, after all. can't expect miracles for that price.

 

[Magic Carpet]

https://bitcointalk.org/index.php?topic=1433925.msg15438155#msg15438155

it's ok, guys

just a minor setback. it's a 200$ card, after all. can't expect miracles for that price.

How old was the board?

 

[Sweepr]

https://bitcointalk.org/index.php?topic=1433925.msg15438155#msg15438155

Well there are a lot of reports of PCI slots crapping out, I came home to this in my rig that had 3 RX-480s in it. They all went back for a refund today.

Will AMD/AIBs take some responsability here?