This site may earn affiliate commissions from the links on this page. Terms of use.

Radeon-RX480-Feature2

AMD has released an updated driver for its RX 480 GPUs that should permanently resolve the GPU'south power draw consequence. For those of you only tuning in, the brand-new Radeon RX 480 launched last week with an unusual power configuration. Specifically, it drew more current from the PCI Express slot on the motherboard than is allowed by the official PCI-SIG specification. While the chances of a trouble or failure were low, AMD promised an updated driver that would resolve these issues permanently. The company released Catalyst 16.7.one yesterday afternoon, and nosotros've put the driver through its paces.

What the new driver changes

There are two changes beingness introduced in the 16.7.ane driver (the launch commuter for the RX 480 was Catalyst sixteen.six.two for those of you keeping score at home). First, AMD has changed how the GPU draws ability from the system. Instead of splitting the load roughly fifty/50 between the PCI Express slot and the six-pin power connector, the card now draws more power from the six-pivot slot and less from the PCI Express slot. This modify is baked into the commuter and automatic. Testing from sites similar PC Perspective and THG has confirmed that this shifts roughly 10W of ability over to the PCI Express six-pin connector and reduces the amount of current being pulled off the PCI Express slot. While the six-pin connector is at present operating somewhat out of spec, it's also designed to handle 8-9 amps per pin and is overengineered by a considerable margin.

22-Overview-Current

Image by Tom's Hardware

Tom'due south Hardware's testing shows that the amount of current pulled from the mainboard has fallen from 6.7 amps to 5.nine amps. While this is much closer to the v.5 amp specification defined past the PCI-SIG, it doesn't pull the GPU 100% into compliance. That'southward where the driver's other new characteristic comes in: A new (desperately named) "Compatibility Style" that throttles the GPU back slightly to ensure that the chip never exceeds PCI-SIG specifications in any detail. Compatibility mode is off by default in the Catalyst sixteen.7.2 driver; end users who are concerned almost power draw can enable the feature manually if they choose.

Unlike the baseline set up, which has no bear upon on functioning, activating compatibility fashion does accept a small touch on on how the GPU performs. AMD characterized the impact as roughly 3%, but we re-ran our own suite of benchmarks to measure the drib. Some of you were concerned that this modify would significantly affect RX 480 performance — just that's non what we saw.

RX480Retest-Perf

Nosotros've gone back and included additional decimal places to ostend that the changes (or lack thereof) aren't being subconscious by rounding. Compatibility mode operation is on top, followed by Catalyst 16.vii.1 without compatibility fashion enabled, followed by the 16.6.two launch driver.

Equally you can see, the degree of difference beyond the three drivers is extremely pocket-sized. The RX 480 is slightly faster in Total War: Rome 2, tied in Ashes of the Singularity, tied in Dragon Historic period Inquisition, slightly faster (five.v%) in Shadow of Mordor, roughly 5% slower in compatibility style when playing COH 2, very slightly slower in BioShock Infinite (less than 2%), and the aforementioned speed, for all intents and purposes, in Metro Terminal Light Redux. Overall, that works out to some small gains without Compatibility Mode enabled and apartment performance with it. Frame timing and smoothness are also unaffected.

Finally, we tested total power consumption at the wall. While we can't measure out the private ability runway, we tin check to see how the GPU is behaving as a whole.

RX480-PowerNewDriver

In other words: AMD appears to accept delivered precisely what it said it would. Full GPU power consumption drops very slightly with the compatibility fashion enabled and increases very slightly with the new driver otherwise, but doesn't shift significantly in either direction.

Overclocking and undervolting

There accept been reports that undervolting the RX 480 can ameliorate its performance, so I took the GPU out and tested it to see what I could observe. In theory, the adaptive voltage and frequency scaling (AVFS) that AMD uses for the RX 480 should have captured most of the headroom available to the chip before — and that seems to exist the case.

Undervolting1

Undervolting our card simply modestly improved performance.

In his article at LegitReviews, Nathan Kirsch discusses how dropping the voltage on his RX 480 from 1100mV and 1137mV at the highest ability states improved performance by roughly three%. Our sample GPU uses tighter voltages — just 1068mV at State half dozen and 1106mV at State 7. While these differences might seem also modest to matter, power consumption in GPUs rises roughly as the cube of the voltage — pregnant small changes can yield significant differences. In our case, undervolting the GPU to 1050mV and 1090mV only improved performance by less than 2% (from 47.33 to 48.2 FPS). Giving the board an extra 50% power headroom didn't make a departure in overall performance.

While our chip's voltage may be lower than some of the cards on the market, that doesn't seem to interpret into better overclocking headroom. Our chip wasn't stable much above 1300MHz, with a maximum frame rate of 49 FPS (upwards from 47.33 FPS) in Metro Concluding Light Redux. We're not really comfy pushing the RX 480 faster at this betoken, mostly because it's not clear if doing and then will result in pulling more electric current from the PCI Express slot than information technology is specced to handle. AIB cards from AMD's partner vendors with more robust power circuitry could unlock more headroom, but for now it seems wise not to push this detail envelope — especially when some users are seeing equivalent performance gains and better ability consumption just from undervolting the card. We've asked AMD for additional information on how the RX 480 will handle overclocking power distribution and volition update this if we hear dorsum.

Final thoughts

We're glad to see AMD responding to this situation speedily and decisively, with a driver update that solves the problem without impacting overall operation. The 16.vii.ane driver doesn't change any of our conclusions about the RX 480 — this GPU is still vastly more power efficient than any previous $200 menu from AMD, and it offers a great combination of performance and features at its price point.

Only — and I really tin can't stress this enough — this is now the third major launch in a row that AMD has self-sabotaged through a series of completely unforced errors. Back in 2022, the R9 290 and 290X were tarnished by poorly designed, noisy reference coolers that struggled to keep both GPUs beneath 95C. Afterward designs from AMD's board partners proved that the R9 290 and 290X could be cooled more effectively and quietly, merely also late — people looked at the reviews of the reference models and thought the GPUs were clinking furnaces.

The R9 Fury X terminal year may not have been noisy, but it however had a noise trouble, thanks to $.25 of glue that caused a peculiarly annoying high-pitched whining unrelated to the usual problem of curl whine on high-stop cards. Again, AMD fixed the consequence relatively quickly, but information technology all the same fed perceptions that the visitor wasn't firing on all thrusters.

Last year, AMD reorganized its unabridged graphics segmentation under Raja Koduri, launche and set out to rebrand its new RTG (Radeon Applied science Group) as distinct and dissimilar from what had come up before. Nonetheless hither we are again.

Problems like this don't simply feed negative perceptions of AMD. Worse, they obscure the very real improvements the company has made to its commuter software, its driver cycle release time, frame pacing, and underlying hardware. When third-political party vendors release their own custom RX 480 designs, in that location's going to be an ongoing discussion about whether or not their cards avert this problem and, if so, how they avoid it. A relatively minor flaw will end up embedded in future reviews because someone signed off on a configuration that shouldn't have shipped, potentially compromising how consumer's view the product in the process.

2016 isn't over. With Vega launching later on this year, RTG still has an opportunity to demonstrate that it can build and launch an uncompromisingly great GPU with no errata or issues that require significant mail-launch testing to accost. Here's hoping they practise.