i7 4790K is INSANELY bottlenecking my GTX 1080 ti

[othertomperson]

5 minutes ago, Mick Naughty said:

So now we are making bottle necks? Just dumb to blame a CPU when you play on a baby res and say the card it too good. 

No one's said the card isn't too good, we said that if you play at a very low resolution and unconstrain the framerate then the CPU will be your limiting factor rather than your GPU, hence this 100% load at low GPU utilisation "issue".

 

4 minutes ago, App4that said:

Listen. I admitted I missed the OP never listed the fps they were getting so if the monitor is 60hz theres the GPU utilization issue. 

 

They are saying they're getting 100% CPU utilization in all their games, not a few, all of them. That's NOT your argument that if you remove the monitor the 4790k can be the limiting factor. You're being an ass, frankly. The OP has an issue, and rather than seeing that you're pushing an off topic argument. 

IT'S NOT OFF TOPIC IT IS LITERALLY OP'S "PROBLEM".

 

In all of those games at 1080p the 1080 Ti is not remotely coming close to its maximum framerate before the 4790k is. You are not finding the 4790k a bottleneck because you are capping fps below its capability. OP isn't, and so they are hitting the maximum that CPU can give, and as a result are sitting at 100% load while the GPU has little to do.

 

I don't get why you're so hung up on the monitor's refresh rate. That video I posted before of getting ~130 fps in Rise of the Tomb Raider was being played on a 60 hz display. The monitor's refresh rate has absolutely nothing to do with what in-game performance you can get.

[Catalonia]

On 4/20/2017 at 2:43 AM, i_build_nanosuits said:

that's the kind of monitors that the GTX 1080ti should be used with:

https://pcpartpicker.com/product/WDcMnQ/acer-monitor-umcx1aa002

 

or at least...this:

https://pcpartpicker.com/product/scM323/aoc-ag271qg-270-165hz-monitor-ag271qg

You forgot one of the best ones:

Asus PG279Q IPS G-Sync 165Hz 1440p

 

 

[App4that]

Time stamp 12:54... Please stop.

 

 

[othertomperson]

5 minutes ago, App4that said:

Time stamp 12:54... Please stop.

I give up. I give up you can keep believing it's fucking magic.

 

OP, just turn v-sync on or cap to your monitors refresh rate and ignore it, it's not an issue.

[App4that]

Just now, othertomperson said:

I give up. I give up you can keep believing it's fucking magic.

 

OP, just turn v-sync on and ignore it, it's not an issue.

Run out of insults? Or realized that even at 1080p games are not held back by a 4790k? 

[othertomperson]

1 minute ago, App4that said:

Run out of insults? Or realized that even at 1080p games are not held back by a 4790k? 

Look there are only two possibilities: 1) you're a troll. 2) you're never going to get it. So why bother either way?

 

Look at the document that goes with that Gamers Nexus review. Read their testing methodology and find out from them why they test at 1080p, and how it's precisely because the GPU stops mattering at that resolution and brings the CPU's limitation to the forefront.

[App4that]

Just now, othertomperson said:

Look there are only two possibilities: 1) you're a troll. 2) you're never going to get it. So why bother either way?

You say troll yet you are the one ignoring facts. I've shown evidence, and admitted when I was wrong. Show me a 4790k holding a 1080ti back at 1080p ultra setting and I will admit it again. Or keep throwing insults without anything to back them so everyone knows who's, who. 

[othertomperson]

2 minutes ago, App4that said:

You say troll yet you are the one ignoring facts. I've shown evidence, and admitted when I was wrong. Show me a 4790k holding a 1080ti back at 1080p ultra setting and I will admit it again. Or keep throwing insults without anything to back them so everyone knows who's, who. 

You already did that yourself. You showed me a load of CPUs tested in the same game at the same settings at the same resolution on the same GPU all getting different framerates. They were all bottlenecking, but because of their differing performance that bottleneck was at a different framerate for all of them. Like I said before, go and read GN's testing methodology.

 

All of the quad cores in that list would have been pegged at 100%. The six to eight cores not so much, they would have been limited by the game engine's ability to scale by cores, but the quad cores were all exhibiting the same behaviour OP described.

[App4that]

Just now, othertomperson said:

You already did that yourself. You showed me a load of CPUs tested in the same game at the same settings at the same resolution on the same GPU all getting different framerates. They were all bottlenecking, but because of their differing performance that bottleneck was at a different framerate for all of them. Like I said before, go and read GN's testing methodology.

2 fps is bottlenecking LOL!!! Hold on, I'll post the 720p scores. 

 

Still waiting on your evidence.

[App4that]

7 minutes ago, othertomperson said:

You already did that yourself. You showed me a load of CPUs tested in the same game at the same settings at the same resolution on the same GPU all getting different framerates. They were all bottlenecking, but because of their differing performance that bottleneck was at a different framerate for all of them. Like I said before, go and read GN's testing methodology.

This should do it, 4 min mark. All the CPU's score the same. Obviously the 1080ti is the limitation. 

 

 

I know someone has a test at 720p, just can't remember who did it. That had fps in the hundreds. 

[othertomperson]

6 minutes ago, App4that said:

This should do it, 4 min mark. All the CPU's score the same. Obviously the 1080ti is the limitation. 

 

 

I know someone has a test at 720p, just can't remember who did it. That had fps in the hundreds. 

All this demonstrates is that in one particular game the 7700k and 1800X are equivalent when paired with a given RAM speed. If you bothered to continue watching after that point you'd see other games that did show one CPU performing better than the other. If Hardware Unboxed did their testing correctly then all of these tests are CPU bottlenecked.

 

I mean holy crap how do you explain an 800 MHz overclock in the 7700k increasing fps by around 30 in GTA V in that video if it was never limited by CPU in the first place!?

 

This is the article that goes with your GN video in which they show the full suite of benchmarks. 

 

http://www.gamersnexus.net/hwreviews/2822-amd-ryzen-r7-1800x-review-premiere-blender-fps-benchmarks/page-7

 

There are two components that are variables here. One is at 70%, the other maxed out at 100%. My argument is that the one that is at 100% is at the maximum framerate it is capable of. Your argument is that that component is fine, it's the one at 70% that's at its limit, and you're wondering why there's an imprint of my palm on my face.

[App4that]

Just now, othertomperson said:

All this demonstrates is that in one particular game the 7700k and 1800X are equivalent when paired with a given RAM speed. If you bothered to continue watching after that point you'd see other games that did show one CPU performing better than the other. If Hardware Unboxed did their testing correctly then all of these tests are CPU bottlenecked.

 

This is the article that goes with your GN video in which they show the full suite of benchmarks. 

 

http://www.gamersnexus.net/hwreviews/2822-amd-ryzen-r7-1800x-review-premiere-blender-fps-benchmarks/page-7

 

There are two components that are variables here. One is at 70%, the other maxed out at 100%. My argument is that the one that is at 100% is at the maximum framerate it is capable of. Your argument is that that component is fine, it's the one at 70% that's at its limit, and you're wondering why there's an imprint of my palm on my face.

No... I have to head to class, please read.

 

The OP states in all their games the CPU sits at 100%, that's what is being focused on as not normal. If you want name a few games and I'll run them at lower resolutions, I've tried GTA in the past and the scaling goes to shit. 

 

I can run Sniper Elite 4, or any other game in my library and show you the utilization. Honestly I usually see it's the scaling that limits me, or RAM speed. But if it will help get the point across I'm williing to try just about anything at this point. 

 

Not for you, could give a rats ass if you want to sit in the corner convinced a 1080ti has no dancing partner, or shouldn't run at 1080p. My concern here is the OP that obviously has an outside issue. And getting that issue resolved. 

[othertomperson]

6 minutes ago, App4that said:

No... I have to head to class, please read.

 

The OP states in all their games the CPU sits at 100%, that's what is being focused on as not normal. If you want name a few games and I'll run them at lower resolutions, I've tried GTA in the past and the scaling goes to shit. 

 

I can run Sniper Elite 4, or any other game in my library and show you the utilization. Honestly I usually see it's the scaling that limits me, or RAM speed. But if it will help get the point across I'm williing to try just about anything at this point. 

 

Not for you, could give a rats ass if you want to sit in the corner convinced a 1080ti has no dancing partner, or shouldn't run at 1080p. My concern here is the OP that obviously has an outside issue. And getting that issue resolved. 

It's not that the 1080 Ti doesn't have "a dancing partner", but it's being used for the wrong purpose. The maximum FPS of the 1080 Ti far outsrips the maximum fps of the 4790k, and in fact any CPU currently available. This is why it is used to test CPUs at this resolution. Usain Bolt may be the fastest sprinter alive, but when he's walking a grandmother across the street he is not going to display his full capabilities. He's "bottlenecked" by the grandmother.

 

Quality benchmarking relies on this being true. The GN results you presented have an entire section on their testing methodology which covers this. Failing that, look at DigitalFoundry because they do the exact same thing for the exact same reason. Or PCper, they also do it. Any competent hardware analyst uses this principle.

[App4that]

2 hours ago, othertomperson said:

It's not that the 1080 Ti doesn't have "a dancing partner", but it's being used for the wrong purpose. The maximum FPS of the 1080 Ti far outsrips the maximum fps of the 4790k, and in fact any CPU currently available. This is why it is used to test CPUs at this resolution. Usain Bolt may be the fastest sprinter alive, but when he's walking a grandmother across the street he is not going to display his full capabilities. He's "bottlenecked" by the grandmother.

 

Quality benchmarking relies on this being true. The GN results you presented have an entire section on their testing methodology which covers this. Failing that, look at DigitalFoundry because they do the exact same thing for the exact same reason. Or PCper, they also do it. Any competent hardware analyst uses this principle.

But the OP is reporting a problem within the limitation on the monitor, which the 4790k CAN drive. No one is saying if the 1080ti could do 500fps the 4790k could, the issue is we are talking (most likely) below 144fps, which is in the range of the 4790k!

 

I keep saying you're off topic, because you are.

[xAcid9]

100% CPU usage in all games? I don't think that possible especially in Overwatch unless you have some malwares running in the background.

[disenter]

Interesting... I'm getting a Strix 1080Ti OC in the next couple of days. I also have a 4790k, but I do have a Asus PG279Q as well. I figured the CPU maybe limit it slightly, it is 3 years old after all, but wouldn't expect it to be to bad. 

I also moved from a GTX 980 to this, so see how it goes and let you know... 

[disenter]

Right, well, my 1080Ti came earlier than expected....

 

so a few benchmark results with my i7 4790k..

 

3dmark firestrike

Gtx980 - 12000

1080Ti - 20000

 

3dmark firestrike ultra 

1080Ti scored just under 8000

this was about 2000 better then what it rated as a "4k gaming machine" and better than 82% of other machines tested.

 

asetto corsa - @ 2560x1440 max details etc it averaged 130fps

@ 1080p it averaged 173fps

 

diablo 3 - 2560x1440 max everything it floated between 120-150 fps, mostly sat around 140. When I entered a dungeon it floated between 250-270 fps.

 

mass effect: andromeda @ 2560x1440 max everything, sat around 120fps most the time, a few areas it got down to 99fps but that's about it...

 

so far very happy with it, and if this is bottlenecked by the 4790k, well, I think it can stay that way a little while... fairly convinced that just about any game I throw at it is going to be easily "playable" and then some! 

[epiqpwnage]

On 4/20/2017 at 8:10 AM, Max_Settings said:

Hello, I just received my Aorus GTX 1080 ti to replace my GTX 980. I have been testing it out in games and in every game I have tried so far I have noticed very low GPU usage. In all the games I have noticed that my i7 4790 overclocked to 4.7ghz was at 100% usage. I have trad Overwatch, Watch Dogs, Ghost Recons Wildlands, and a few more games, and in every one my CPU is at 100% but my GPU is only sitting around 70% usage. I am playing at 1080p so maybe that's the issue? Is there a way to get more FPS out of this card?

Your cpu is bottlenecking because 1080ti is too fast @1080p)and the cpu can't keep up. Try playing on 1440p or 4k or even an ultrawide 1080 monitor and the bottlenck goes away

[Max_Settings]

21 minutes ago, epiqpwnage said:

Your cpu is bottlenecking because 1080ti is too fast @1080p)and the cpu can't keep up. Try playing on 1440p or 4k or even an ultrawide 1080 monitor and the bottlenck goes away

Yes I have solved my issue. But why did you comment on a post back from April? Like why were you looking so far back in posts? 

[Fir33]

On 24.07.2017 at 2:03 AM, Max_Settings said:

Yes I have solved my issue. But why did you comment on a post back from April? Like why were you looking so far back in posts? 

What was the problem in your setup ?

 

Asking because i'm planning to buy new GPU and 240 hz monitor and i have 4790k 4.6 which got 80 % load by 1060 3gb G1 Overclocked  when i play Overwatch with uncapped fps and lowest settings .

 

 

 

 

[ov3rk1ll]

I know that this is an old thread, but I have a similar setup to OP, with no bottle-necking.

 

I'm using 3x 1080p monitors @ 60hz, 4790k @ 4.6ghz(HT Enabled), MSI GTX 1080ti Gaming X(+150mhz core clock, +350mhz Memory clock with Afterburner). 32GB Corsair Vengeance(4x8GB modules) @ 2400mhz .

 

I disable v-sync in my games and enable Fast Sync in Nvidia control panel to remove tearing(Fast Sync does not work for DX12, Vulkan, or OpenGL, so I either enable vsync for those games or use adaptive sync).

 

My CPU usage only hits 100% in 2 games(Wolfenstein:The New Order and Hitman 2016), but my 4790k has no problems with other CPU bound titles like GTA V and Crysis 3. Generally my CPU usage in games is 55%-85% on both single/triple screen modes. I'll admit that my GPU usage doesn't always hit 100% in games, but it is never held back by my 4790k as far as I can tell.

 

I tested Wolfenstien:The New Order and Hitman 2016 with a GTX 960 4GB in my setup and still hit 100% CPU usage, so those a badly optimized games-not a bottle necking issue.

 

 

[Mick Naughty]

Technically it's still a bottle neck in that term regardless of software. It's still happening so it's can be called anything. 

[ov3rk1ll]

@Mick Naughty- I get what you're saying Mick, it is technically a bottle neck(even though I said my system has no bottle necks), but I should clarify: an often overlooked factor is bad optimization. The bottle neck isn't caused by mismatched hardware, or playing at 1080p on a 4k GPU, the bottleneck is caused by the way those games utilize your CPU at max settings.Hell, I got bottle necked using a 1080p GPU(GTX 960).

 

Maybe these games require more cores, but honestly if a 4790k can't handle a game, I would point my finger at the developer, rather than my hardware.

[Mick Naughty]

That's how bf1 was in release.  4790k ans sinlge gtx 1080 on 1440p. 100% use. Still have issues today with my spare rig on 1080 on old ass hardware.

[ov3rk1ll]

Thanks for the heads up, I haven't gotten around to playing BF1(mostly because storm of steel is so breathtaking, that I keep replaying that level).

 

Just realized that I used terms that I don't fully agree with in my last post-4k GPU and 1080p GPU. I keep seeing people commenting on how a GPU is meant for a specific resolution and using a monitor that is of a lower resolution will degrade performance and bottle neck the system, but the truth is your performance will not be degraded by an overkill GPU. It may not max out the GPU usage, but this is actually a good thing, because it has the added benefit of lower temps and power usage which could theoretically increase the GPU's lifespan.

 

I actually upgraded from a GTX 1070, because contrary to what everybody says, this so called "2k GPU" is not capable of running certain games at 1920x1080p maxed out(Rise of The Tomb Raider springs to mind) without dipping to less than 40 fps in most instances. When I upgraded to my "4k" GPU, I was able to play those same games at 80-120 fps without needing to lower my graphics settings. My 1070 could barely run GR:Wildlands at 1920x1080p on medium settings, now I can run it maxed out with a 1080ti on the same hardware at 80+ FPS with no issues.