Nvidia Pascal cards still exhibit high-refresh-rate “power bug”

A while back, we reported on a weird issue with GeForce graphics cards and multiple-display setups with refresh rates higher than 60 Hz. The short version is that connecting multiple high-refresh-rate displays to the DisplayPorts on a GeForce card causes it to run at much higher clock speeds than the Windows desktop would seem to demand, increasing power draw and heat production. As Pascal cards start to trickle into the TR labs, we figured it was as good a time as any to revisit this issue and see whether it's been ironed out. Sadly, the problem appears to persist with these next-gen cards.

First, a brief word about our test setup. The particular graphics card we're using is Gigabyte's GeForce GTX 1080 Xtreme Gaming. My test bench has two monitors on it: my personal Eizo Foris FS2735, a 144-Hz FreeSync display, and Asus' ROG Swift PG279Q, a 165-Hz G-Sync display. Both of these screens have DisplayPort inputs, and I hooked them up to the GTX 1080's corresponding DisplayPort outputs. I monitored clock speeds using Gigabyte's Xtreme Gaming Engine software.

Our GTX 1080 running at idle, as it should be

To reproduce this bug, I first tested the card with the PG279Q connected by itself. Whether the display was set to run at 60Hz or 165Hz at the Windows desktop, the GTX 1080's core clock speeds fell to the expected 291MHz at idle. Going by our trusty Watts Up power meter, the system consumed about 65W to 75W on average. Setting the PG279Q to run at 165Hz and plugging in the Eizo at 60 Hz caused the card to clock up to 1304 MHz. At those speeds, our system draws about 107W on average. Setting the Eizo display to run at 144Hz didn't cause clock speeds or power consumption to increase any further.

Once again, though, things can get weird with this setup. Running the Eizo display at 144 Hz with the PG279Q set at 60 Hz causes the GTX 1080 to clock down to 291 MHz at idle, as expected. As we saw above, though, the opposite isn't true. Disabling G-Sync on the PG279Q doesn't cause the GTX 1080 to clock down, either. I'm at a loss to explain why any of those behaviors are as they are.

Our GTX 1080, reproducing this "power bug"

Thanks to Pascal's apparent efficiency gains, the average power draw of our system with the GTX 1080 clocked up is a pretty significant reduction from the 130W to 150W our test system pulled from the wall with a similarly-afflicted GeForce GTX 980 Ti installed. Still, it's clear that Pascal can run even more efficiently yet when its clock speeds aren't being pushed above what's apparently needed to drive the Windows desktop. Users with multiple high-refresh-rate displays are still going to pay more for power over the life of the card than they otherwise might if the card clocked down appropriately at idle, and it appears this problem resides somewhere in the driver.

For now, I'm resorting to setting both of my displays to run at 60 Hz at the Windows desktop so that I'm not inflating my power bill and unduly heating up my office. I'm still hoping Nvidia finds it in its heart to investigate and fix this issue, because fiddling with Windows display settings every time I want to take full advantage of the PG279Q's 165-Hz refresh rate is far from ideal. We'll keep you updated if we hear anything more about this issue from the green team's HQ.

Comments closed
    • End User
    • 3 years ago

    My setup:

    1) Primary monitor (XB271HU) is connected to the 1080
    2) Primary monitor is at 150 Hz when idle on the Windows desktop
    3) Secondary monitor connected to motherboard (iGPU)
    3) The 1080 idles at 139 MHz / .625 V / 36 °C

    • Pendragon80
    • 3 years ago

    Oh having slightly higher clocks is preferable to AMD’s choice of annoying flickering on desktop or amd grey screen bug, go gripe about that.

    • Pendragon80
    • 3 years ago

    So strange that this so called “bug” is brought up because it’s not a bug.Nvidia cards have worked this way since Kepler and this “bug” won’t change.

    1.Set your main display to 120hz in cp.
    2.Set highest available refresh rate in cp.

    This will keep clocks low on desktop and automatically switch to fastest refresh rate in apps.The “Bug” is fixed.

    • Phartindust
    • 3 years ago

    After reading the article and comments, I’m wondering is this an issue only in setups with monitors that have different refresh rates, or does it occur also with matched monitors?

    Also, if you have matched monitors, does the bug appear if you change the refresh of just one of them?

    • ultima_trev
    • 3 years ago

    What’s the problem? 😛 NVidia clearly just wants its customers to enjoy extreme 165Hz desktop productivity performance.

    • Rza79
    • 3 years ago

    These guys already reported on it a month ago:
    [url<]http://www.computerbase.de/2016-05/geforce-gtx-1080-test/10/#diagramm-leistungsaufnahme-des-gesamtsystems-windows-desktop[/url<] Power usage also seems to jump with three monitors attached: [url<]http://www.computerbase.de/2016-05/geforce-gtx-1080-test/10/#diagramm-leistungsaufnahme-des-gesamtsystems-multi-monitor-betrieb[/url<]

    • Klimax
    • 3 years ago

    Again? Simply doesn’t make sense anymore. Why doesn’t NVidia’s testlab catch this? And frankly, why wasn’t support correctly integrated into driver. Looks like they still keep it hacked in…

    Not good.

    • khanmein
    • 3 years ago

    try to turn off MFAA on global setting, “The stub received bad data” will randomly pop out. y didn’t u guys report this issue??

    • Leader952
    • 3 years ago

    [quote<] I'm still hoping Nvidia finds it in its heart to investigate and fix this issue[/quote<] Did you report your problem to Nvidia?

    • moriz
    • 3 years ago

    hmm, my reference GTX 1080 idles at 1164 MHz core with three monitors attached (73 Hz and 2x 60 Hz). i thought it was standard behavior.

    speaking of “power bugs”, my GTX 1080 would randomly have its TDP limit set to a low number, like 40%. this would cause the card to downclock dramatically while gaming. i managed to fix (or mitigate) this issue by installing EVGA Precision and manually setting the TDP limit to 120%.

    i’d be interested to know if anyone has had a similar issue.

    • Chrispy_
    • 3 years ago

    Typical Nvida; They fix something that has needed fixing for years, then break it shortly afterwards.

    I’m living with it on my 970 right now by running at 60Hz (My screens are only capable of 85Hz anyway) but this is yet another stupid bug to add to the decade long list of things that Nvidia can’t be arsed to fix. It shows just how disjointed their driver team(s) are and how little they work together.

    AMD are not without their issues, but I’m not keeping a mental list of stupid issues and workarounds for AMD like I’ve always had to with Nvidia; Once a bug is fixed for AMD, it usually stays that way.

    • DoomGuy64
    • 3 years ago

    Hey guys, wanna hear about some AMD bugs?

    1: 390 since doom update no longer downclocks memory speeds / uses low power memory states on the desktop. Clockspeeds are normal however. This behavior can cause flickering in web browsers possibly due to ram artifacting. (fans don’t spin up unless GPU gets hot.) 1b: Clock throttling on some models. Don’t really care for the power efficiency “fix”, because it’s not a real answer.

    2: All sorts of flickering with Fury series. Latest driver claims to fix it.

    3: Tonga series: All of the above. They seem to be hit the hardest.

    4: Multimon breaks things. Most complaints come from multimon users.

    [url<]https://www.reddit.com/r/Amd/comments/4ky4tz/amd_stahp_every_driver_after_1642_has_this/[/url<] [url<]https://www.reddit.com/r/AMDHelp/comments/4osjsz/r9_380_aknowledged_by_amd_crashing_problems_hasnt/[/url<] [url<]https://www.reddit.com/r/Amd/comments/4niv4p/psa_furyfury_x_users_that_suffer_from_display/[/url<] For me, drivers were fine up until the Doom update, which was necessary to play the game. Solved my memory clock issue with afterburner's automatic clock management settings. Still waiting for official fix. Plenty of issues on AMD side ATM, so nvidia users are not alone.

      • Vaughn
      • 3 years ago

      Why do you feel its necessary to talk about AMD driver issue when topic is clearly about Nvidia?

      Defensive much?

      I applaud those that have kept it on topic and giving their own experience with your nv cards.

        • NoOne ButMe
        • 3 years ago

        Indeed. They both have issues. All users should know this.

        The only difference is AMD tends to have more issues unresolved and Nvidia appears to have an issue keeping the problems staying fixed.

        • DoomGuy64
        • 3 years ago

        Mostly balance and fairness, because there isn’t just a singular driver issue. Amd has some issues that almost rival this one from Nvidia, and their issues are worse than power use. Screen flickering and clock throttling render the card unusable for people experiencing it, and it doesn’t seem like amd is taking these problems serious. Not to mention amd also had their own issues with 144hz, as well as multimon. But hey, feel perfectly free to never discuss it at all.

      • Kougar
      • 3 years ago

      No, I want to hear about NVIDIA fixing this bug already. It’s old, it’s been acknowledged, and NVIDIA claimed it got patched at one point. The longer it takes NVIDIA to resolve this the more it seems like a hardware issue, because if it was a fast, simple software fix then why didn’t they do it months ago.

        • NoOne ButMe
        • 3 years ago

        [STARTSARCASM]it is arriving with the Async driver for Maxwell and the Dx12 driver for Fermi[/ENDSARCASM]

      • Klimax
      • 3 years ago

      Frankly, keep this thread on topic. Also one might argue that AMD bugs are not really of much interest or common. Just 20% common…

    • Andrew Lauritzen
    • 3 years ago

    I still get this occasionally on my 980 as well – it’s definitely indicative of some complex set of power heuristics at play that tend to fail sometimes. Display-port’s way of “sleeping” displays (especially Gsync ones) seems to exacerbate issues as well, as the display entirely disappears from the OS and comes back, after which sometimes things are wonky. Behavior I get randomly spans the range here:

    1) Sometimes it reverts the display to 60Hz.
    2) Sometimes I get the high power use thing at 144Hz. It does seem to happen less frequently at 120Hz but I’ve seen it happen there as well.
    3) Occasionally I get a stranger issue where the GPU settles at 500Mhz and *doesn’t clock up* even in games resulting in ~1/2 performance or less.

    A reboot almost always fixes all of the above, but it’s slightly annoying in this day and age where there’s almost no need to reboot outside of those kinds of issues (I don’t think I’ve rebooted my work machine for months and it has zero issues).

    • Voldenuit
    • 3 years ago

    [quote=”Jeff Kampman”<]For now, I'm resorting to setting both of my displays to run at 60 Hz at the Windows desktop so that I'm not inflating my power bill and unduly heating up my office. [/quote<] @Jeff, I would at least experiment with 120Hz windows desktop to see if that works for you. Scales nicely with 24p, 30p and 60p video. And the main reason I don't run 144 Hz desktop, aside from the power bug.

      • ColeLT1
      • 3 years ago

      Agreed, 120hz on the desktop should fix it. I have my gsync set to 120 (and my 2nd monitor is a 60hz). Games still run up to 144, and the card does reach idle.

      • Jeff Kampman
      • 3 years ago

      120Hz lessens the issue but it doesn’t solve it. The card still wants to hang out in the 800MHz range.

        • ColeLT1
        • 3 years ago

        Is that with both monitors on 120Hz? Each setup is going to be different but mine is 120Hz and 60Hz and the card goes all the way down. Also, I’m on a GTX970 so that matters, but I have a 1070 on the way.

        • Voldenuit
        • 3 years ago

        I notice also that you are using Gigabyte software. I found that Gigabyte Easy Tune would be a culprit in pegging GPU and CPU utilization. Once Gigabyte nixed fan controls in Easy Tune, I ditched the rest of their software.

        msi Afterburner seems to be working well for me.

          • ColeLT1
          • 3 years ago

          My 1070 will be in Monday and I will check in. I also will be only using afterburner.

    • xeridea
    • 3 years ago

    This silly bug should never have existed. Clocks should in no way be tied to refresh rate, it should be entirely based on GPU load, which at desktop should be about 0.1% for a high end card.

    The bug went unpatched for ages before (IIRC it was known about for about a year before it became a hot topic, then it took a few driver releases before they “fixed” it). Now, several months and drivers later, on a new line of cards, the issue still exists. Any efficiency gains are pretty much nullified by the cards using a lot more power for hours on end while idle.

      • GTVic
      • 3 years ago

      It is not necessarily tied. The card uses a set of conditions to determine the clock speed, the high refresh rate may trigger some part of the card to turn on which in turn triggers something else etc. etc. The conditions may be extremely complex.

        • xeridea
        • 3 years ago

        The same parts of the card would be on for 30Hz or 165Hz. The utilization of the card would still be negligible. If the card can’t do 144Hz without burdening some part of it, that would be a poor design. It isn’t tied to resolution, so 4k @60 should be more taxing than 1080p@144. We have phones with crazy high resolutions using 100x less power, it is crazy to think the card couldn’t handle it without high clocks.

        • ColeLT1
        • 3 years ago

        I have to agree with this, because they have proven that it is possible to have the low clocks at 144Hz+ with previous drivers that were “fixed.” Then they removed the fix, I bet it caused a situation, maybe a race condition or something, where the driver or card would have a failure event in certain situations.

    • rjohnson11
    • 3 years ago

    Are you using the hotfix driver?

    [url<]http://nvidia.custhelp.com/app/answers/detail/a_id/4166/%7E/geforce-hotfix-driver-368.51[/url<] best regards Robert rjohnson11 (a.k.a rjohnson11, EVGA volunteer forum moderator)

      • Jeff Kampman
      • 3 years ago

      I gave it a shot, but the issue persists even with the hotfix driver. Thanks for bringing it to my attention, though.

    • Voldenuit
    • 3 years ago

    I have a GTX 970 w an Asus PG278Q (144 Hz G-Sync) and an acer KHUL27bmiidp (60 Hz AHVA). GPU clocks idle down normally at 120 Hz desktop on the Asus (135 MHz), but will clock up to 600-800 MHz at 144 Hz. I run 120 Hz on the desktop, and it behaves well with most of my games, but a few games will reset my desktop to 144 Hz after exiting, and I will have to manually re-select the refresh rate.

    PS: Some games (mostly newer ones) let you select the in-game refresh rate, eg on Overwatch, I have a range of options @1440p from 60Hz-144Hz. You can also make per-application profiles for refresh rate and G-Sync support in the nvidia control panel, but that can be a bit of legwork.

    • OneShotOneKill
    • 3 years ago

    These cards suck.
    Go AMD Wooohooo!

    Oh wait, AMD cards probably draw that much just to display the windows logo loading.

      • OneShotOneKill
      • 3 years ago

      Downvote already!

        • Meadows
        • 3 years ago

        Mom’s spaghetti!

          • MarkG509
          • 3 years ago

          Don’t forget the [b<]cheese[/b<]!

            • chubbyhorse
            • 3 years ago

            FTFY
            Don’t forgetti*

    • chuckula
    • 3 years ago

    [quote<]First, a brief word about our test setup. The particular graphics card we're using is Gigabyte's GeForce GTX 1080 Xtreme Gaming. [/quote<] So..... possible review of a non-FE card from TR coming in the future?

      • Captain Ned
      • 3 years ago

      Jeff said in another thread that it’s due this week.

        • blahsaysblah
        • 3 years ago

        Nooo too early. I request delay for adding i3-6320 vs i5 benchmarks for sweet spot.

        edit: Oops. GTX 1070, not GTX 1080 review.

Pin It on Pinterest

Share This