GTX 1080 Overclocking Results Thread - Post yours here!

kring

Limp Gawd
Joined
Aug 27, 2004
Messages
165
I didn't see another thread started on OC results for GTX 1080s, so I thought I would start one. the intention is just to post results and not necessarily to discuss.

EVGA Founders Edition
120% Power, 92c Temp, linked with Power as priority
Running quiet fan curve

Stable Results:
+180 GPU (Instability at 200)
+360 MEM (Artifacts at 400)

I backed it down to 150/330 to ensure headroom, at these settings It hits peek boost at 2050 but settle into games at 1974 on GPU and 5335 on MEM and running 82c temp. it idles at 31c temp, for some reason though the fan is always running at 1100 even though the fan curve shows it should be off. I think it's a driver/precisionX bug.
 
Last edited:
I didn't see another thread started on OC results for GTX 1080s, so I thought I would start one. the intention is just to post results and not necessarily to discuss.

EVGA Founders Edition
120% Power, 92c Temp, linked with Power as priority

Stable Results:
+180 GPU (Instability at 200)
+360 MEM (Artifacts at 400)

I backed it down to 150/330 to ensure headroom, at these settings It hits peek boost at 2050 but settle into games at 1974 on GPU and 5335 on MEM and running 82c temp.


Are you running default fan curve? i see anyone running around 70-75% fan keeps temps under 75c.
 
If anyone would stock these damn things, I would have something to share. :\
 
I am running the quiet fan curve. good Q, I'll add that to the original post.
 
I am running the quiet fan curve. good Q, I'll add that to the original post.
You might get a more stable, higher overclock if you turn the fan speed up with a custom curve. I'd start by setting the max to 70-75% at 80C and go from there, depending on your noise tolerance.
 
+190 / +400 is about all I can manage. Driver crashes at 200 and thats with 100% fan curve starting at 80C.

Oddly enough I don't get any increase in score on Valley nor Heaven benchmarks from stock to this OC though. Blaming it on the fact that the new SLI HB bridges arent out yet though.
 
+190 / +400 is about all I can manage. Driver crashes at 200 and thats with 100% fan curve starting at 80C.

Oddly enough I don't get any increase in score on Valley nor Heaven benchmarks from stock to this OC though. Blaming it on the fact that the new SLI HB bridges arent out yet though.

Update, most games worked fine but Overwatch locked up at this setting. Finally landed on +150 / +350.
 
You might get a more stable, higher overclock if you turn the fan speed up with a custom curve. I'd start by setting the max to 70-75% at 80C and go from there, depending on your noise tolerance.

Thx, tried that. got 2012 and 5355 with temps in the high 70's... still couldn't get much higher without artifacting or crashing. plenty happy with that for now. this is to just hold me over till I get EVGA classified version when it's available.
 
Not really going all out, but good with +90 and +400. Just used the settings on Guru3D for Afterburner. No desire to go any higher as It easily maxes out all of my 1080p games. The extra nudge doesn't matter in the 4K games I can't get 60fps in, so another 1-2% won't matter any more. I'm happy.
 
I haven't done any overclock yet, but I did adjust my fan profile using the Afterburner 4.3.0 Beta 4. At stock settings the fan was capping around 50-55% which, while quiet, seemed to hit the 80-82 degree ceiling relatively quickly when playing GTA V maxed out at 4K resolutions (minus AA, since 4K on a 27" does a pretty good job at masking aliasing). This would cause the boost clock to vary quite a bit, constantly going up and down and sometime dipping below the max stock boost clock (1733 MHz) but I don't recall ever seeing it dip below stock clock (1607 MHz). By changing the fan profile to run the fan at around 50% at around 60 degrees, 75% at 80 degrees, and 100% for everything else above my boost clock has done a much better job at settling to a fairly fixed rate and stays pretty consistently above the stock boost clock with no other changes. My boost clock during extended plays in GTA V so far appear to stay over 1800 MHz, settling around 70-100 MHz over stock boost clock for the duration of game play. Being in a fairly SFF setup, Corsair Obsidian 250D, I don't think that is too bad considering no other adjustments were made.

I do plan on tinkering with the fan profiles a little more before I decide to work on the clock speed of the GPU and memory. I may have it go to 75% sooner and may have it increase slightly higher at 80 degrees. I know at 75% you can hear the airflow but it isn't annoying nor does it wash out the sound, but it is noisy at 100%. I usually play my games with headphones on, but I want to be able to tolerate the sound if I choose to use my speakers, thus my reason for not going straight to 100%. The silver lining, at least, is all the noise is airflow and not fan whine. I don't hear the fan at all, just a whooshing noise. And while the case has ventilation in the side panel next to the video card, there is a gap between it and the video card, so I think the video card is still sucking in some warmer air from inside the case. I may try to devise a "gasket" of some sort around the video card fan that presses up against the side panel vents so as to increase the amount of outside, cooler, air the video card pulls in. Any thoughts or opinions on this?

I'm also actively looking for a larger, higher flow fan to replace the case's sole intake fan on the front that is quiet. The case can take up to a 200mm case fan, but due to the H100i V2 AIO I have installed, I am going to be limited to fans that are no deeper than 25mm. Currently I am temporarily using a Thermalright TY-140 that I swapped out the stock Corsair AF140L for. The Thermalright pushes just a hair more air (73.6 CFM versus 67.4 CFM), but I mainly swapped out the Corsair for this because it is quieter and is PWM. The Corsair had a bit of a whine. So something that is just as quiet (or quieter) that pushes at least as much air as the Thermalright, I'd be happy with.

On a different note, I think I may have stumbled across a bug in the latest Afterburner beta. When I initially installed the software, I didn't make any changes to the card. The only changes I made to the software was to have it output CPU and GPU stats to the LCD on my Logitech keyboard. But when I went to fire up a couple of games to benchmark that GTX 1080 wouldn't even ramp up to stock speeds, it would sit at around 1066 MHz and go no higher. Interestingly, this put the performance of my GTX 1080 at about the same level I was seeing with my MSI GTX 970 Gaming 100ME 4G. Luckily, the fix was easy: just reboot the PC. After reboot, the GTX 1080's clock speeds began behaving normal again and the performance in my games jumped back up to where they were before Afterburner was installed.
 
Has anyone tested the minimum fan percent required to maintain 2Ghz on a FE without throttling? Seen lots of vague reports you need to sit at 70% inside a well-ventilated case. Is 70% loud?
 
While my case is several feet away, it's sitting right beside my PS4. The PS4's fan is easily louder than my 1080's fan blaring at 80+%. Apples to apples, it isn't as loud as my 980 running at the same speed either. It's probably only 2/3 as loud. Running it at 60-70% all of the time wouldn't be horrible.
In my 10-15 minutes of testing, I didn't get throttled at 60%, but I was just a hair under 2GHZ.
 
Doesn't the default profile run at 55% at max temperature? Another 5% fanspeed was enough to avoid throttling?
 
Doesn't the default profile run at 55% at max temperature? Another 5% fanspeed was enough to avoid throttling?

Not sure - I just set the fanspeed manually in Afterburner based on what some others were suggesting on Guru. At the default fan setting, my 3DMark total was 500+ points lower, so clearly it was throttling.
I'll need to experiment a bit more over the weekend. I really only had a few minutes to test some basic things out. The dynamic fan profiles might also be of use, too.
 
Without OC my card sits around 1875 MHz in game, with +200 it gets up to 2050 MHz while benching. Temps are just fine with a pretty standard fan profile in MSI Afterburner. I have never experienced throttling. I've been running reference NVIDIA since the olden days so I know that with a blower card you need to tune the fan to crank up at certain temps. Mine will go 100% fan once 80C is reached. I typically see around 65C while in game (Ark).

I guess the stuttering and throttling is a big deal to some...but it can be easily fixed with a new BIOS or a new driver. In my case, MSI Afterburner.

Some of the reviews and things that I have seen are so different than my experience I start to wonder if these folks have bad cards or are purposely trying to post FUD due to hate for the FE? I mean, I get it. It is a "rip-off" at $100+ but there's no way I'm running dual or tri-fan cards with SLI today and the always possible addition of a third 1080 for PhysX.
 
At 2050mhz on Core. 5500Mhz on Memory. Constant 75% fan speed and never goes above 80 degrees even under load.

Fire Strike score:

Sgs2mau.jpg
 
at 1500mhz my 980 ti scored a 21800 graphics score. I only have a 4 core processor so cant really compare physics, but I am really surprised to see a 2050 1080ti only outscore my 980ti by like 8%

y3m0Bws_D3SHRIWmB1bWLimkk_YV0TuVJ5b-fs7OOrMLVKRxbUgi1RaNYwQIJ8CiUgHmdKCX8iCuKr4TUNWWhnLhmH-SiMa0FK0V-4uD8I_SM_SZJH0KLsD6Bpn7zrMznJy581C7f5J3tdQ1U2IsDCLkGdQBbwH97gFYd00q5JJi88
 
My Firestrike score was roughly 17,600 at stock speed and 18,200'ish with my (mild - +90 and +400) OC.
 
at 1500mhz my 980 ti scored a 21800 graphics score. I only have a 4 core processor so cant really compare physics, but I am really surprised to see a 2050 1080ti only outscore my 980ti by like 8%

The 1080 is not that much faster than 980 Ti / TITAN X in 3DMark. No surprise there.
 
Using Guru3D's mild overclock settings to start off with (+90 and +400), my initial Fire Strike score was only 12,844. Something clearly wasn't right. I ran the benchmark multiple times with similar results. I started to compare the data to others with similar setups and noticed my frame rate was lower than theirs. Seeing that I realized what the problem was, I had G-Sync turned on since I have a G-Sync monitor. So despite an application or game having vsync off, G-Sync will come in and skew your results. G-Sync is great for gaming but bad for benchmarks. So I turned G-Sync off and reran the benchmark and got a score of 17,984. That looked much more in line with what others are getting with similar setups.

My highest overboost clock speed spike at 1,987 MHz and settled in around the high 1,800 to low 1,900 MHz range.

On an another note, Afterburner beta gimped my clock speeds again after fiddling with some fan profile adjustments. This time the GPU wouldn't want to go any higher than 895 MHz. Card was barely breaking a sweat and the fan saw no need to ramp up. Rebooting the PC solved the issue. Another problem I saw was when I enabled core voltage offset and Afterburner restarted (as it stated it needed to), it disabled my fan profiles without me knowing. So my first run I noticed halfway through that the fan was going no higher than 50-55%, hampering my boost clocks. At least that was quick to detect. But just an FYI for others out there using the beta, keep an eye out for similar issues.
 
Using Guru3D's mild overclock settings to start off with (+90 and +400), my initial Fire Strike score was only 12,844. Something clearly wasn't right. I ran the benchmark multiple times with similar results. I started to compare the data to others with similar setups and noticed my frame rate was lower than theirs. Seeing that I realized what the problem was, I had G-Sync turned on since I have a G-Sync monitor. So despite an application or game having vsync off, G-Sync will come in and skew your results. G-Sync is great for gaming but bad for benchmarks. So I turned G-Sync off and reran the benchmark and got a score of 17,984. That looked much more in line with what others are getting with similar setups.

My highest overboost clock speed spike at 1,987 MHz and settled in around the high 1,800 to low 1,900 MHz range.

On an another note, Afterburner beta gimped my clock speeds again after fiddling with some fan profile adjustments. This time the GPU wouldn't want to go any higher than 895 MHz. Card was barely breaking a sweat and the fan saw no need to ramp up. Rebooting the PC solved the issue. Another problem I saw was when I enabled core voltage offset and Afterburner restarted (as it stated it needed to), it disabled my fan profiles without me knowing. So my first run I noticed halfway through that the fan was going no higher than 50-55%, hampering my boost clocks. At least that was quick to detect. But just an FYI for others out there using the beta, keep an eye out for similar issues.

895MHz? Sound like you had a driver crash. When my 980 Ti has a driver crash, the 3D clocks stick at 495MHz until I reboot, and your 895MHz would follow since the 1080 has a much higher clock speed.
 
Guru3d just reviewed the MSI 1080 Gaming X (non-reference board with an 8 pin and 6 pin power connector) and it kind of overclocks like shit. Not really any different than the founders edition.

MSI GeForce GTX 1080 GAMING X 8G review

I read that all cards are voltage locked so it makes sense that they are all equal regardless of cooling solution installed.
ASUS ROG STRIX GeForce GTX 1080 offers poor overclocking | VideoCardz.com

That said ComputerBase confirmed the rumors from PCGamesHardware. The so-called Extreme overclocking (2050+ MHz) on STRIX GTX 1080 is poor. All cards are locked to 1.25V maximum, and when the voltage gets close this value, card become unstable. Neither bypassing VRM controller nor removing TDP limit has made any difference. However once LN2 is in action, the clock frequency can go up to 2400 MHz.

Either way, the chase of mysterious 2.114 GHz clock advertised by NVIDIA has still not found its conclusion, even on this custom card.
 
895MHz? Sound like you had a driver crash. When my 980 Ti has a driver crash, the 3D clocks stick at 495MHz until I reboot, and your 895MHz would follow since the 1080 has a much higher clock speed.
It's happened once before but the GPU clock wouldn't go higher than 1066 MHz or so. Both occurred after installing the beta Afterburner. The first time was immediately after installing the software, no changes to the GPU settings, only turned on some of Afterburner's monitoring features to display on the LCD of my keyboard. I didn't reboot the machine after installing the software, but rebooting resolved that problem. The most recent occurred earlier tonight after tinkering with the fan curve profile settings in Afterburner. Again, a reboot solved the problem. If it was a driver crash, the latest drivers and Windows have been pretty good at catching them and recovering the driver without issues. But I certainly do think it has something to do with the beta Afterburner, possibly even a mix between it and the latest Nvidia drivers. With the software still being in beta, I do expect some bugs. Here's hoping they get ironed out before it goes final. I should probably post the bugs on a forum where the developer is watching.
 
It's happened once before but the GPU clock wouldn't go higher than 1066 MHz or so. Both occurred after installing the beta Afterburner. The first time was immediately after installing the software, no changes to the GPU settings, only turned on some of Afterburner's monitoring features to display on the LCD of my keyboard. I didn't reboot the machine after installing the software, but rebooting resolved that problem. The most recent occurred earlier tonight after tinkering with the fan curve profile settings in Afterburner. Again, a reboot solved the problem. If it was a driver crash, the latest drivers and Windows have been pretty good at catching them and recovering the driver without issues. But I certainly do think it has something to do with the beta Afterburner, possibly even a mix between it and the latest Nvidia drivers. With the software still being in beta, I do expect some bugs. Here's hoping they get ironed out before it goes final. I should probably post the bugs on a forum where the developer is watching.

FWIW I haven't experienced any of this with my usage of 4.3.0 Beta 3 of Afterburner...
 
FWIW I haven't experienced any of this with my usage of 4.3.0 Beta 3 of Afterburner...
I'm using 4.3.0 beta 4.

Might be a specific set of conditions? Other items that are tied to the graphics card in one shape or form is my 4K G-sync monitor with 4k enabled and my Spyder3 Elite software. The computer build is fairly new, I built it at the end of April/beginning of May. I haven't received any BSODs nor do I see any errors in my Event Logs.
 
I'm using 4.3.0 beta 4.

Might be a specific set of conditions? Other items that are tied to the graphics card in one shape or form is my 4K G-sync monitor with 4k enabled and my Spyder3 Elite software. The computer build is fairly new, I built it at the end of April/beginning of May. I haven't received any BSODs nor do I see any errors in my Event Logs.

I'm using G-Sync, too - just 1440P though (Acer Predator). Thanks for the heads up on the new build - I'll try it and see how it goes. I recall that when I had 680s I would regularly have to reboot my system when the driver would be "stuck" I.E. low MHz or very little to no FPS. A reboot would fix it every time. This is when I was doing NV Surround with 3x1080p/120Hz monitors. I suspected it was more NV Surround than anything else.
 
Ok I'm at +260 , +500 in sli with EVGA cards and testing, it will throttle down to around 2100 and back up in games.
 
Ok I'm at +260 , +500 in sli with EVGA cards and testing, it will throttle down to around 2100 and back up in games.


Make sure ECC on the memory does not kick in at 500mhz and you get lower performance, reading around the forums ECC kicks in around 350-450mhz so the best bet is to have it at 300mhz max for the Memory Overclock.
500mhz might look impressive but means nothing if you actually loose performance.
 
OK finally got my 1080, and went from

21805 graphics to 24091. also about 1000 higher than Zamardii's 1080 score. Not too shabby. about a 11% increase over my 1500mhz 980ti


1080gtx overclocked.jpg
 
OK finally got my 1080, and went from

21805 graphics to 24091. also about 1000 higher than Zamardii's 1080 score. Not too shabby. about a 11% increase over my 1500mhz 980ti

redir

11% is a pretty weak upgrade though.
 
11% is a pretty weak upgrade though.

True, but this is 3dmark and not even extreme or ultra. I understand games respond much better to the 1080. I actually bought a 980ti like 20 days before the 1080 was announced. I returned it for a refund, and have been videocardless minus my old 6950 ati card lol. finally got the 1080 in the mail today, so i think i lucked out.
 
EVGA Founders Edition: +200/550Mhz, Afterburner 100% power, 120% voltage. Fan profile modified slightly.
This is inside a mITX NCASE M1. I have a Noctua fan at 900rpm feeding cold air into the GPU intake. GPU consistently stays above 2Ghz on long gaming sessions.

To be honest I am surprised its overclocking so well even with warm ambient temperatures and a small ITX case. I was expecting it to overclock poorly and was just happy to run it at stock speed. I also got lucky with the CPU which is running at 4.7Ghz (6700k). Currently the little rig is #124 in the Top 200 gaming rigs on 3DMark Firestrike I scored 19 573 in Fire Strike





 
Make sure ECC on the memory does not kick in at 500mhz and you get lower performance, reading around the forums ECC kicks in around 350-450mhz so the best bet is to have it at 300mhz max for the Memory Overclock.
500mhz might look impressive but means nothing if you actually loose performance.

I just did some testing on this. First I ran Firestrike at +200/300Mhz. Graphics score: 24 200
Then I immediately did another run on the warm GPU at +200/550Mhz. Graphics score: 24 556

And a run after a reboot with a cold GPU at +200/550Mhz. Graphics score: 24762

So it looks like as far as Firestrike is concerned, you still get better performance at higher Mhz, even if ECC might be kicking in. Or I might have a really good card and ECC is not kicking in at all. No way to tell?
 
Back
Top