AMD Releases Statement On Radeon RX 480 Power Consumption; More Details Tuesday
by Ryan Smith on July 2, 2016 5:05 AM EST
Following up on this week's Radeon RX 480 launch, there has been some questions raised about the power consumption of the card. This is after some sites whom directly tap the power rails feeding the card discovered that at least some of their samples were pulling more than the standard-allowed 75W over the PCIe slot and/or 6-pin PCIe external power connector.
To that end, it would appear that AMD's staff is working weekend duty, and they have just sent over the following statement.
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).
If some of the data is to be believed, these cards are exceeding 150W total at times, which would mean there is either something causing them to run in the wrong power state, or they are just outright exeeding their power limit and need to be throttled back. As we don't do per-rail testing I don't have anything meaningful to add at this second, but it will be very interesting to see how AMD responds next week.
Update 07/06: AMD has since released their status update, which you can find here.
Facebook
Twitter
RSS
181 Comments
View All Comments
euskalzabe - Saturday, July 2, 2016 - link
Yeah it's interesting to see how AMD went for a very cheap reference card, where Nvidia went with higher grade designs. It doesn't make for a good PR comparison in the beginning with both cards being released so close to each other...DanNeely - Saturday, July 2, 2016 - link
With Polaris unable to compete with GP 104 on performance they had to aim for the low price high volume segment of the market. At the same time I suspect the problematic power draw levels from the 480s suggests that they were expecting the reference design to run somewhat lower power, but either didn't quite get the perf/watt they were hoping for or decided to bump clocks when they saw where nVidia's cards were performing.Hopefully they can adjust the relative power draw via a software update to clamp the max drawn through the mobo (the wiring in the PCIe cable is able to safely deliver a good amount of power beyond the nominal 75W rating) and limit the reputational damage. If they can do so the higher price points nVidia's going after with the 1060 should continue to protect AMD from direct competition for a bit longer. OTOH a cut down GP-106 will probably be launching under the GT 1050 brand before much longer and should be able to match the price and performance of the RX480 fairly closely.
Yojimbo - Saturday, July 2, 2016 - link
Polaris on a 14nm process has an efficiency comparable to Maxwell on a 28nm process. I think my statement gives a truer picture of the situation.jwcalla - Saturday, July 2, 2016 - link
I believe that's the real take-away from the 480 release, even more important than this question of PCIe compliance.bill4 - Saturday, July 2, 2016 - link
Maxwell is terrible in DX12, which all future games will use. Also, efficiency doesn' matter, just price and performance. A GPU's job is running games, not being most efficient. In fact efficiency is actually speed vs price, there is no power efficiency that matters, since power draw does not matter to end user.Nvidia got destroyed here, they need to get a Pascal car out quickly in this price range, they've already lost gigantic amount of market share. Unfortunately if rumors are true the 1060 has terrible efficiency, as it's rumored to use a worthless 3GB card at 249, and 299 for the 6GB viable card. As you can see AMD is far more efficient.
Meteor2 - Saturday, July 2, 2016 - link
What rumours? There's nothing to suggest that the 1060 won't have the same excellent efficiency as the 1070 and 1080.DanNeely - Saturday, July 2, 2016 - link
The rumor mill is talking about two different GP106 variants, one with 3gb of ram and the other with 6gb. Some are claiming that the 3gb model will launch as a $250 GT1060 with the 6gb model as a $300 variant. Others that the GT1060 will only come in 6gb models. Presumably in the latter case the 3GB model will be used for the GT1050.If the 6GB model is only available at the $300 price point (and ofc assuming the 3gb one ends up bottlenecked by less VRAM), it'd be 50% more expensive than the RX480 for only about 15% -25% more performance (according to what are alleged to be leaked nVidia slides). If that's the case the GT1060 would be seriously overpriced for the relative performance gain it offers.
I'm hoping that they'll all be 6gb cards and the $250/300 price spread is custom cooled vs founders edition variants. This seems most likely to me; $250 for a 3GB card seems seriously overpriced today, when the 950 was available in a 4gb config for well under $200 a year ago.
Meteor2 - Sunday, July 3, 2016 - link
I thought we were talking about efficiency, not value? I.e. the ability of the 480 to do what it promised within its power budget -- in particular, within the spec of the power supply connections.MapRef41N93W - Saturday, July 2, 2016 - link
Yep end users don't care about power draw. It's why the GTX 750ti was the most bought card in the world from it's release till sometime this year. Oh wait....Idiot.
Shodoman - Sunday, July 3, 2016 - link
Oh, i don't think it has much to do with power draw. Most users don't even know what that is. You're forgetting how most components/computers are acquired.Unfortunately with your last word you proved you are not civilized enough to have a conversation. Which also undermines your arguments somewhat. Pity...