• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

[Various] NVIDIA GeForce GTX 1080 Review Thread

Page 32 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
My 1080 doesn't throttle below boost clocks. You can keep looping over and over. But retail and the reviews dont even use the same driver. The fan issue that .25 got isn't in the beta drivers the reviews used either.

So why dont you show me a card with .25 driver that throttles? I already proved mine doesn't go below boost clocks, even in the worst possible scenario with stopped case fan and high ambient.

The difference between you and me is I actually own the card and know how it behaves. Its quite clear you have no interest in Nvidia cards besides trying to downplay them as much as possible.

Great since your 1080 doesn't throttle you can make a YouTube video showing how it doesn't throttle. Just show your fan curve then let a game warm the system up for 20 minutes at load then record the Temps/clockspeed at 100% gpu usage so we can see how well the 1080 holds up and whether they fixed the throttling issue.

Please don't make this personal let's stay on topic. This is about the 1080 and not about my own personal preferences. It's an unfair attack to say I have no interest in Nvidia products.

Either way though, like I said, 1080s may throttle based on the substandard cooler. If you have low enough ambient Temps or a golden chip of course you may be one of the lucky few with a non throttling chip. Just like there were people with fury x with out pump noise. Doesn't change the fact many people had a substandard experience.

It's a shame that people are willing to ignore Nvidia's substandard cooling situation this time around. We need to have high expectations of Nvidia if we want them to continue operating at a high level.
 
We had this discussion in another thread, seems the performance increases are pretty much exactly 30% for each new release. Meaning that the "big chip" is not "better" in any way. The small chip makes up for size and cores, with better process and architecture. The end result is the same. The 980Ti was a fine chip, but it wasn't "special" and nor will the 1080Ti be. Being on the "small chip" cycle is no worse than being on the "big chip" cycle. In fact, the small ones seem to be priced slightly lower (although 1080 may have changed that). Nvidia knows where to target the performance, and history suggest they will hit the same exact spot again. It makes no economic sense to release something that has a bigger performance jump unless AMD trounces them which seems unlikely in the short term.
We did and when I agreed with the sentiment it was under the assumption that the 1070 and 1080 had the same oc headroom maxwell did and the same scaling.

When I think back I realize how naive that was since pascal has a lower IPC than maxwell. So adding oc to the 1080 doesn't put it as far past stock clocks as it did on the 980.

Because of this the 1070fe is a slower 980ti aib and the 1080fe is 5-15-% faster due to throttling. Aib 1080s and when Nvidia decided to make better oc support more available and we'll see if the 1080 has any real legs to stand on calling itself a 980 successor.

If the 1080 is going up against an improved perf/watt vega chip, I feel bad for it. This is less impressive than maxwell launch.
 
Last edited:
http://www.computerbase.de/videos/2016-05/nvidia-geforce-gtx-1080-lautstaerke-im-test/

1080 reference card sounds like a hair drier, just like the 980ti, 980 etc. This is nothing new though, just the $100 extra for the 'premium components'. It's the only reason the card gets mocked; if reference remained at MSRP like in the past no reviewers would have said anything.

The miracles of Pascal; 'premium components', faster than 980 sli, great overclock, low temps etc. were as empty as most claims of miracles are.

Founders was a marketing failure. An otherwise decent release had a turd dropped on it. You can't call a card premium when it's louder, throttles, overclocks poorly, runs hot and is not as over designed as cheaper AIB cards.
 
Last edited:
Ok big guy. I'll start a thread asking for 1080FE owner participation. Wont be til tonight though.

Still waiting for any sources claiming it is fixed.

Here are more people with issues though: https://www.reddit.com/r/pcmasterrace/comments/4mlfme/_/

...

Anyway, today I decided to return these cards because the thermals are awful. I was expecting maybe a bit of throttling if I OC'd the cards like crazy and didn't run maximum fan, but what I actually got was unbelievable.

...

TL;DR: The GTX 1080 FEs run far too hot to actually use their power.

P.S.: Before anyone mentions it, I knew the cards throttled a bit when I purchased them, I just didn't know they throttled to this extent. The reference coolers have always been bad in comparison to AIB coolers, but these are just in another league of bad. I've learned my lesson, won't buy reference ever again.
 
http://www.computerbase.de/videos/2016-05/nvidia-geforce-gtx-1080-lautstaerke-im-test/

1080 reference card sounds like a hair drier, just like the 980ti, 980 etc. This is nothing new though, just the $100 extra for the 'premium components'. It's the only reason the card gets mocked; if reference remained at MSRP like in the past no reviewers would of said anything.

The miracles of Pascal; 'premium components', faster than 980 sli, great overclock, low temps etc. were as empty as most claims of miracles are.

Founders was a marketing failure. An otherwise decent release had a turd dropped on it. You can't call a card premium when it's louder, throttles, overclocks poorly, runs hot and is not as over designed as cheaper AIB cards.

I disagree. I think it was a marketing success. They sold a throttling card (at least for a significant subset of users - doesn't seem to be every card) with a crappy reference cooler for $100 more than custom cards. They took advantage of people who couldn't wait and I'm sure it'll help my NVDA holdings when they announce sales numbers.
 
I wouldn't buy a 1080TI, because I already got a 1080. And assume the 1080TI is 250W or so. That's higher than I wish to have.

So 599 to 699 is an awful price. Yet people happily paid 649 or more for 980TI, Fury X and Nano? 🙂

The double standards are massive, I know 😀

The only reason why there is so much fuss about this, is because one IHV didn't bring anything to the fight and those people have to wait till 2017 for whatever that comes at whatever it may be priced.
How many people happily paid for 980ti's and fury X's that had a sub par cooler at an inflated price?

People buying those cards got either an awesome aftermarket cooler or a liquid cooled card. Fury nano was a niche card for ultra small system builders which is why it commanded its price.

Stop defending the 1080 cooler. People are attacking the cooler and it's price hike, not the actual gp104 chip.

Sent from my XT1575 using Tapatalk
 
Crazy how the only people complaining are the ones who don't own the card. How often do you see that with retail products? D:
 
Crazy how the only people complaining are the ones who don't own the card. How often do you see that with retail products? D:

Except we have reports of people complaining about throttling already. Hence how we know about the throttling because people experienced throttling. Read above to see some of those complaints.
 
http://www.computerbase.de/videos/2016-05/nvidia-geforce-gtx-1080-lautstaerke-im-test/

1080 reference card sounds like a hair drier, just like the 980ti, 980 etc. This is nothing new though, just the $100 extra for the 'premium components'. It's the only reason the card gets mocked; if reference remained at MSRP like in the past no reviewers would have said anything.

The miracles of Pascal; 'premium components', faster than 980 sli, great overclock, low temps etc. were as empty as most claims of miracles are.

Founders was a marketing failure. An otherwise decent release had a turd dropped on it. You can't call a card premium when it's louder, throttles, overclocks poorly, runs hot and is not as over designed as cheaper AIB cards.

It's only a marketing failure if it doesn't sell.

So far, it's sold out in most places.

Now, I won't defend such a product going for a premium, but I have to give kudos to NV's marketing team, they found a way to extract even more profits from consumers by selling them an inferior product for higher prices.
 
I don't know how smart it is from their marketing, they got plenty of negative press from it. Amazon reviews are like at 1 star.
 
I don't know how smart it is from their marketing, they got plenty of negative press from it. Amazon reviews are like at 1 star.

It's pretty smart. They know that the majority of sites will test in an open test bench so can set the fan profile fairly low for initial reviews. This way they get performance and a good rating on the noise tests too. Once reviews are in no sites are going to back and re-review just for a driver change so those values (noise especially) are kind of set in stone and will be referred back to when comparing cards in the future.

Once reviews are done and sealed then release an updated fan profile for the consumers that keeps the card cooler and helps the throttling issue at the expense of fan noise.
 
Only reason the amazon reviews are so low is because of the people selling it at 800-1000 dollars originally and any time they are sold out.

As far as the fan noise, has anyone heard the FE at 100% fan speed. It's not loud....
 
Only reason the amazon reviews are so low is because of the people selling it at 800-1000 dollars originally and any time they are sold out.

As far as the fan noise, has anyone heard the FE at 100% fan speed. It's not loud....
why not share it in this thread? 😀 would put to rest alot of the arguments.

I got curious, and googled for a noise video. https://www.youtube.com/watch?v=fbHfuovpjFk it is basically the same as a 980 n 980 ti reference. where are the noise complains coming from?
 
Last edited:
It's a shame that people are willing to ignore Nvidia's substandard cooling situation this time around.

Ive always wondered about this myth. Since when did the Titan cooler become substandard? Have you ever used an AMD blower before? especially when its used on a GPU that consumes 50~80W less than the GTX980Ti. Did you know that the GTX980Ti, GTX980, 780 etc all do throttle?

My GTX780 uses the exact same cooler with a different fan curve (it is at high % based on temp because to my surprise these coolers dont have that annoying whine tone to it that most blowers do unless your running it at 100%). Playing Doom for instance, 76~78C at my overclock and you can just hear the air whooshing a little bit.

Actually im beginning to think that the word throttling is inaccurate because your not actually downclocking from the base reference. The baseclocks are always guaranteed (stock clock) and boost clocks are basically giving you extra performance given power/temp limits. So in fact your increasing your stock clocks if there is power/temp headroom. Now if it went below baseclocks that would actually mean it is throttling. The boost clocks has and always has been "YMMV" type feature.

I find it amusing the amount of downplaying of this card. The price yes could be argued for, but theres so much hate and personal emotions involved. Grooveriding for instance (not a call out by the way) is even saying its loud - https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1080/25.html. Titan cooler at 55% fan speed isn't loud.

Don't really see the fuss at all. Blower type coolers have their applications but for the majority of us AIBs will give us what we want at a lower cost. It takes time for these models to start hitting the market. Not the end of the world folks.. plus aren't most of you guys not even looking for 1080 type performance cards anyway?

Edit - The VC&G forums have become so toxic lately with so much brand vs brand fighting. Can't have proper discussions. Many new members are here to just join in with the mud slinging, and times where posters shared their setups, overclocks etc seem like a rarity these days or the talk about games/graphics. Its ALL about AMD the righteous vs nVIDIA the wrong.
 
Last edited:
It is loud. I posted a video. Personal investment is common here, but it's generally the posters who make claims without evidence because they need them to be true, or if that won't do, try and get the false claims to be perceived as true.

In this case the discussion is noise, so an example of the noise needs to be given. I gave a video evidencing the card is loud even at low fan speeds. Imagine 100%... actually I don't need to having owned quite a few cards with this similar blower on them. They've all been loud. The reference blower is a loud card.

Edit - The VC&G forums have become so toxic lately with so much brand vs brand fighting. Can't have proper discussions. Many new members are here to just join in with the mud slinging, and times where posters shared their setups, overclocks etc seem like a rarity these days or the talk about games/graphics. Its ALL about AMD the righteous vs nVIDIA the wrong.

This adds nothing to the discussion. We are discussing tangible and measurable metrics. This is your perceived subjective assessment of what you think posters here are thinking in their minds while they post. This could easily be reversed to righteous nvidia and AMD the wrong. Sticking to the subject would be better and actually add to the discussion. Guesswork as to other posters' perceived or real bias adds nothing.
 
Last edited:
I find it amusing the amount of downplaying of this card. The price yes could be argued for, but theres so much hate and personal emotions involved. Grooveriding for instance (not a call out by the way) is even saying its loud - https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1080/25.html. Titan cooler at 55% fan speed isn't loud.

An example..

JayZ's review: open bench, 80% fan speed to prevent thermal throttling.

https://www.youtube.com/watch?v=OXUo1S55ZUM

80% fan speed is no longer considered quiet. Only apologists would say such things.

gtx-1080-review-noise-levels.png


Certainly the blower on the 290X was horribly noisy on auto settings, at 49 dB. A reference 1080 at 100% fanspeed is 57 dB.
 
It is loud. I posted a video. Personal investment is common here, but it's generally the posters who make claims without evidence because they need them to be true, or if that won't do, try and get the false claims to be perceived as true.

In this case the discussion is noise, so an example of the noise needs to be given. I gave a video evidencing the card is loud even at low fan speeds. Imagine 100%... actually I don't need to having owned quite a few cards with this similar blower on them. They've all been loud. The reference blower is a loud card.

Sensitive ears you have! I can barely hear 55% unless I put my ears close to the case. Well guess it differs between people. I find 60~65% acceptable for me. Anything beyond that it gets quite loud. 100% is definitely loud.

But I want to mention that for me it doesn't have that whiny tone (high pitch) to it that most blowers do, especially the one AMD use to use for 3~4 generations. This has been the most annoying part of coolers for me. That such pitch whiny noise when the RPM gets too high. However for the titan cooler, it just sounds like air whooshing so its ok for me.
 
So what you are saying is the 1080 at 100% fan speed on the FE is quieter then the gtx 980, the gtx 980ti vr, and of course the amd r9 290x ref. So in reality, the gtx1080 isn't loud. Gotcha.

Especially when the aib coolers come out, everyone will realize that these aren't loud cards. However I guess everyone has a difference of opinion to loud and it depends on there case, the placement of it etc. To me, 57DB inside of a case, isn't loud at least.

Sound is subjective to everyone, but this just proves that at 100% fan speed, it's quieter than the 980 and 980ti at 100% fan speed.
 
An example..

JayZ's review: open bench, 80% fan speed to prevent thermal throttling.

https://www.youtube.com/watch?v=OXUo1S55ZUM

80% fan speed is no longer considered quiet. Only apologists would say such things.

gtx-1080-review-noise-levels.png


Certainly the blower on the 290X was horribly noisy on auto settings, at 49 dB. A reference 1080 at 100% fanspeed is 57 dB.

But like i mentioned, this behavior has been seen with all previous nVIDIA cards. I really don't think its a problem. Maybe the high expectations? fair enough seeing as it is a 16nm product plus $100 more expensive. For instance, even hardOCP mentions this for their GTX780Ti review -http://www.hardocp.com/article/2014/01/13/asus_r9_290x_directcu_ii_oc_overclocking_review/3#.V1YwjbhcRBc.

Boostclocks is like an auto overclocking feature based on temp/power. You should know this. Obviously better coolers (axial over radial) and better power delivery circuit will have better boost characteristics.

I would hazard a guess that a better custom fan curve (that doesn't require it to go to 80%) would dramatically increase the reference boost clock profile for the GTX1080 like what I saw with my 780.
 
An example..

JayZ's review: open bench, 80% fan speed to prevent thermal throttling.

https://www.youtube.com/watch?v=OXUo1S55ZUM

80% fan speed is no longer considered quiet. Only apologists would say such things.

gtx-1080-review-noise-levels.png


Certainly the blower on the 290X was horribly noisy on auto settings, at 49 dB. A reference 1080 at 100% fanspeed is 57 dB.
80% is like a hair dryer alright, but it is what it needs to not throttle.

so hot n noisy? 😀 I guess it is.
 
Last edited:
But like i mentioned, this behavior has been seen with all previous nVIDIA cards. I really don't think its a problem. Maybe the high expectations? fair enough seeing as it is a 16nm product plus $100 more expensive. For instance, even hardOCP mentions this for their GTX780Ti review -http://www.hardocp.com/article/2014/01/13/asus_r9_290x_directcu_ii_oc_overclocking_review/3#.V1YwjbhcRBc.

Boostclocks is like an auto overclocking feature based on temp/power. You should know this. Obviously better coolers (axial over radial) and better power delivery circuit will have better boost characteristics.

I would hazard a guess that a better custom fan curve (that doesn't require it to go to 80%) would dramatically increase the reference boost clock profile for the GTX1080 like what I saw with my 780.

Oh I am well aware, except the reference 1080 has taken it to a whole new level.

Compare the boost clocks within the first few minutes of a benchmark, I mean JayZ's video demonstrate it boosting to 1.87 to 1.89ghz.

After awhile it drops to 1.62ghz. The delta is incredible, 250 to 270mhz! Much more than 780Ti, 980 and 980Ti ever behaved.

This calls into question the accuracy of tech press benchmarks.

1. If they do open bench, it skews towards higher boost clocks as compared to gaming inside a case.

2. If they do short benchmarks, it heavily skews towards the higher boost clocks.

Imagine you as a buyer, read these sites, you see 25-30% faster than 980Ti numbers, yet when you take it home, put it into a case, after 10 minutes of gameplay, it's now only 15% faster than a 980Ti because it's running at close to it's base clocks.

To prevent that performance loss, you have to crank up your fan, at which point, bye bye quiet.

I mean most of us already knew custom cards would be heaps better, but with further testing and these review re-visits, it's clearly not a premium build at all going for a premium $.
 
Once again and again... nvidia without a doubt, no little twinkle of disbelief has the absolute fastest card. And once again, capitalism apologists aside, no one is ok with a chip half the size costing 40% more for 15% more performance.

Why is it that people can't admit that they wanted the performance of the 1080 but think it should have cost less? If you were just hanging out with nvidia lovers would you then admit it in secrecy? Nvidia does not operate a fab nor incur r&d costs for new nodes. That's TSMC's job and early adopters like Apple helped pave the way to finance this development on 16nm.

Nvidia may pay an incremental amount higher per wafer for 16nm for now, but are they paying 280% more per die mm? (1080 is 40% more than a $500 980ti, yet half the die size). If you tell me you believe that then you are either trolling or truly blindfolded.

TSMC did not triple their costs to customers especially with competition heating up from people like GloFo and samsung.

Its amazing the lack of compromise I see in these forums. People are so far "left or right" that they absolutely refuse to offer concessions in debate.

Nvidia can't adjust the fan speeds or anything to do with hardware you need a Bios update not a driver update because there will be many different fans also profiles soon with AIB's partners.
 
Last edited:
Back
Top