AMD RX 480 Power Issue Detailed - VRMs Beefier Than GTX 1080 Founder's Edition

The AMD Radeon RX 480 has come under scrutiny since its launch 2 days ago due to controversy surrounding PCIe power draw of some GPU samples sent to the press. Several reviewers take found that their RX 480 review units drew more than 75 watts of power from the PCI Express slot. Which according to the PCIe spec, the slot should non provide more than 75 watts of power.

AMD RX 480 PCIe Slot Power Draw Issue Detailed

PC Perspective, a well respected engineering science publication, has since published its results upon farther investigating the issue. Ryan Shrout, Editor-In-Chief at PCPer.com, conducted PCIe slot & 6-pivot ability draw testing with an elaborate setup that interrupts any electric current going through the PCIe slot and 6-pin ability connector before information technology goes to the card.

To mimic a realistic workload the ability measurements were taken in The Witcher three & Ascent Of The Tomb Raider. In Rise Of The Tomb Raider the load on the PCIe slot'southward twelve volt ability rail hovered effectually seventy watts throughout with a couple of spikes going upwards to 77 watts.

Credit : PCPer.com

In The Witcher 3 the results were virtually identical. Power depict from the motherboard's 12 volt runway hovered effectually ~73 watts throughout with two, less than one second, spikes going up to 77 watts. Clearly then on both occasions the RX 480 exceeded the 75 watts. In both occasions however it was marginal at all-time, lasting for less than one second.

AMD's RX 480 Isn't The First Graphics Card To Exhibit Such Beliefs - Custom GTX 960s Caused Similar Concerns Terminal Twelvemonth

Bug at the Motherboard Slot

We've got to go back to the foundations article mentioned above to put the measurements at separate rails into context. This is because the otherwise very adept Asus GTX 960 Strix leaves the motherboard connector to bargain with unprecedented unfiltered ability spikes all on its own:

The very frequent spikes beyond the motherboard slot'southward supposed limit won't cause firsthand damage to the hardware, but there might well be long-term repercussions that are hard to judge now. The same goes for how the arrangement might otherwise exist impacted with problems such as "chirping" on-lath sound when the mouse is moved. The Asus GTX 960 Strix should do a much meliorate job smoothing these spikes out.

Credit : Tom's Hardware, GTX 960 Review - 2022, January 22nd''

Interestingly, PCPer followed up Tom's testing with their own. Which did not prove a similar behavior on the GTX 960. Although notably, Ryan Shrout explains that this is because a 2nd order filter is applied. To negate the upshot aberrant spikes would have on the data. Which is quite helpful when we want to await at the typical power consumption over a specific catamenia of fourth dimension.

One interesting notation on our data compared to what Tom's Hardware presents – we are using a second order depression laissez passer filter to smooth out the information to arrive more readable and more indicative of how power draw is handled by the components on the PCB. Tom's story reported "maximum" power describe at 300 watts for the RX 480 and while that is technically authentic, those figures correspond instantaneous power draw. That is interesting data in some circumstances, and may actually signal other potential issues with excessively noisy power circuitry, only to united states of america, information technology makes more sense to sample data at a high rate (10 kHz) only to filter it and present information technology more than readable way that improve meshes with the continuous power commitment capabilities of the organization.

Some gamers accept expressed concern over that "maximum" ability draw of 300 watts on the RX 480 that Tom'south Hardware reported. While that power measurement is technically authentic, it doesn't represent the continuous power describe of the hardware. Instead, that measure is a result of a high frequency information acquisition system that may take a reading at the exact moment that a power stage on the card switches. Any DC switching power supply that is riding shut to a sure power level is going to exceed that on the leading edges of phase switches for some minute amount of time. This is another reason why our low pass filter on ability data tin help represent real-globe power consumption accurately. That doesn't mean the spikes they measure are non a potential cause for concern, that's just not what we are focused on with our testing.

PCPer.com

The event is that when these outliers occur often plenty that they're no longer aberrant. As they're no longer occurring over a minute corporeality of time but are still filtered out, essentially reducing the reliability of the data.

Looking at Tom's Hardware'southward RX 480 PCIe slot power measurements. Nosotros meet a higher power draw on average than the GTX 960 only with significantly fewer spikes. Normally spikes can be disregarded every bit the issue of an electrical component'south switching current. These spikes can and should be ignored, especially when nosotros're talking almost spikes that don't concluding more than a few milliseconds and are a natural phenomena.

All the same, in the case of the GTX 960 we see a spike occur almost every other second that lasts over a substantial enough menses of time to annals on Tom'due south 60 second ability measurement graph. In fact near every i of those spikes lasts for roughly 1 2d and touches 150W. A behavior that is quite dissimilar from the RX 480 and is worth paying attending to.

The principal cause for any electric component failure related to electrical overstress is thermal fatigue. Prolonged, repeated thermal cycling leads to current crowding which then leads to thermal runaway and ultimately component failure.

In other terms the main cause of component failure caused by overcurrent is temperature related. The process of cycling through two temperature extremes is what's called thermal cycling. The higher the current spike - and the more information technology's out of the component spec - the more extreme the delta is between the 2 temperature states. Eventually the oestrus will damage the component and it volition neglect. The bigger these spikes become and the more frequently they occur, the faster this damage will take place.

While GTX 960 ability measurement data sets from Tom'southward Hardawre & PCPer practise show disagreement the publications are in understanding about the RX 480. Their verdict appears unanimous and is this "Don't read also much into it!". As long as the RX 480 is not overclocked, its PCIe slot ability depict should not be cause for business.

PCPer.com - Ryan Shrout
As is usually the case in this market, there are some asking for a recall on the RX 480 or saying that AMD is going to be liable for some kind of grade action lawsuit. I don't agree with either sentiment. AMD is clearly pushing the envelope with the power delivery system on the Radeon RX 480 and we accept shown that they are definitely drawing power outside PCI Limited specifications even at stock clock speeds and power settings. I haven't come across any stability concerns still nor have I heard of whatsoever other reviewers indicating as much. As the RX 480s start to reach consumers hands today and subsequently this calendar week, I volition exist paying close attention to customs reports while doing our own longer term testing at PC Perspective.

AMD Radeon RX 480 Overclocking Limitations investigated

Moving over to overclocked results the story changes. Using the very same Witcher 3 benchmarking run PCPer's overclocked RX 480 drew upwards to 90 watts from the PCIe slot. So clearly this turns from a non-issue to something far more serious. Although Ryan Shrout notes :

Because this occurs only with overclocked settings, AMD is not really responsible for the specification alienation in this specific case, even though the visitor has talked almost the overclocking headroom bachelor on the RX 480 directly.

This is a good reminder as to why GPU vendors, both Nvidia & AMD, are emphatic well-nigh overclocking voiding the warranty. Something that users should keep in heed. It'due south clear then that if y'all intend to overclock you'll be much better served with a custom RX 480 from AMD's AIB partners. Sapphire, XFX, Asus & MSI all have their own custom RX 480 designs lineup for release.

Those are their own RX 480 designs with meliorate cooling and more importantly 8-pivot PCIE power connectors. These tin provide upwardly to 150 watts of power on their ain. Which volition piece of work wonders to alleviate any excessive ability draw from the PCIe slot and improve overclocking. In fact just a couple of days AMD's AIB partners reported that with extremely skilful air coolers they're seeing their RX 480s clock to 1480-1600Mhz.

The reference designed RX 480 already draws 155-165 watts in games. So it'southward at the very limits of what can be delivered through a 6-pin PCIe power connector and the PCIe slot. Merely put it's starved for power for whatever serious overclocking. Expanding the ability headroom past calculation additional 6-pivot or 8-pin PCIe ability connectors or replacing the existing vi-pivot with an 8-pivot should address that straight.

The RX 480's Elaborate 6+1 Phase Design Puts The GTX 1080 Founder's Edition "High Quality" Components Claims To Shame

The VRM power delivery circuitry on the reference RX 480 printed circuit board is beyond impressive. More so when compared to the GTX 1080's five+ane ability phase blueprint which simply pales in comparing. The RX 480's on-lath VRM - voltage regulator - on the high-side is capable of delivering forty amps from each phase at 125c, for a total of 240 amps.

At room temperature each phase tin can evangelize 66 amps for a total of 396 amps on the loftier-side and up to 600 amps on the low-side. The GTX 1080 Founder'southward Edition voltage regulator can evangelize 30 amps at 125c from each of its five phases, for a total of 150 amps on the low side, 246 amps less than the RX 480. At room temperature it goes upwardly to 50 amps per phase for at total of 250 amps, 350 amps less than what the RX 480 VRMs can deliver at the same temperature on the low side. The GTX 1080 actually employs "dr." mosfets which combine the high and low side into one IC. Hence why the PCB can get abroad with one mosfet per phase rather than needing two.

It'southward clear then that information technology's just a affair of feeding the RX 480 with plenty power externally and not a design quality result by whatever ways. Having a practiced base to beginning from gives AMD's AIB partners a swell opportunity to differentiate themselves with their ain designs. Many of which nosotros have seen teased over the past couple of days with more to come.