Really bad game lag with a 1080ti

[LethalWarfare]

I get about 40fps on Medium settings. I should be able to get a lot more than that just like I used to get an avg of 100fps on maxed with the alpha 17 experamental a few months back. Also I am hosting a server with 1 other person on it. The game seems to be limited by the cpu but why tf would I need a cpu clocked at more than 4.3Ghz to run this game? Anyway if anyone knows why this is hapining please respond:)

 

 

PC Specs:

GPU: GeForce GTX 1050 Ti | GeForce GTX 1080 Ti

CPU: AMD Ryzen 7 2700X Eight-Core Processor @ 4175MHz

Memory: 32 GB DDR4 3000MHz RAM

Operating system: Windows 10 Pro 64 bit

 

I have the CPU running at 4.3GHz on 6 cores btw and yes I am using the 1080ti not the 1050ti.

[SylenThunder]

More information is needed, but you should be able to get a pretty stable 60FPS out of it. Average on 100 on a17 will be really pushing it due to lack of client optimization.

 

Also, since you're AMD, you're likely only running on one core, which could be part of your issue.

 

Biggest things, don't run higher than 1920x1080. Turn off shadows and reflections. Run exclusive fullscreen. Oh, and kill all the bloat on your PC. People leave way too many things just running idle and wasting system resources.

[LethalWarfare]

Well its not running on 1 core, in fact it cpu useage goes up on all cores after I start the game, witch is something I hae never seen in any other game. I have tried 1080p but it only boosts my fps by lik2 20-30% and when I get an avg of 25fps during a 21 day hoard. By the time it gets to like the 61 day hoarde I will most likely be getting 10fps. Without shadows the game looks unbareable ugly and unrealistic. This is close to the best gaming PC money can buy right now. I should be able to run at mid settings 1080p at 60fps. There has to be some other reson for the lag then simply my PC isn't capable of running it. IF that was the case then no pc could do 1080p maxed settings.

 

Any other ideas on what might be causeing this?

 

- - - Updated - - -

 

Here is the CPU usage about 45 sec after I started the game. You can see the peek in all cores.

[SylenThunder]

Interesting. The only other thing I can think of is a slow platter drive or system bloat. Or maybe you didn't exclude the client from Antivirus and have one of the especially nasty programs like Norton or Kaspersky.

 

Just for a benchmark comparison, here's what I'm running....

 

Win 10

i7-3930k overclocked at 4.2GHz

16GB DDR3 RAM overclocked at @1600Mhz

GTX1060 6GB with stock OC

OS and game on SSD, save data on ultra-fast HDD's in RAID 0

 

 

On the vanilla game I average 70-95FPS with it more regularly in the 100's. Do have the drops into the 47-60 range when in buildings. A horde night with the zeds-per-wave count at 32 will bring me into the 30's. (Note: day count doesn't mean anything for hordes since a15.)

 

 

I'm running with basically everything on full at 1920x1080 except Reflections, V-Sync, SSOA, DOF, and Sun Shafts are all turned off.

 

 

My system isn't exactly what I'd call clean either, I haven't done a wipe in over a year. It's way overdue.

Basically we're clocking in at the same speeds with the same core count, except you have faster RAM.

[Fox]

wtf? Since when does a 2700x have 6 core 12 threads?

 

Also, my 1600x at 3.9GHz sits between 20% - 30% on near max settings and semi-stable capped 72fps. So something is definitely not right at your end.

 

I do know there's been a lot of complaints relating to 1080ti performance with this game. I'm wondering if maybe it's a heat issue with those cards as Nvidia occasionally skimps out on thermal pads for the memory controllers on their cards.

 

I am curious as to why you would have 2 different graphics cards in your computer though? Maybe there's a conflict and the game is choosing the 1050ti over the other?

 

Also, since you're AMD, you're likely only running on one core, which could be part of your issue.

Not true, this game would likely not work at all on a single core. Also, AMD's version of hyperthreading works with this game now, they fixed it. So now it's basically 2x2 threads spread out on all cores for this game, just like any Intel CPU. It used to be just 2 cores with no hyperthreading, but that got fixed at some point.

[LethalWarfare]

AMD ryzen master allows you yo "disable" cores to decrease heat.

 

- - - Updated - - -

 

Thunder you get the same performance I get on a 1060? How does that work?

[Fox]

AMD ryzen master allows you yo "disable" cores to decrease heat.

 

- - - Updated - - -

 

Thunder you get the same performance I get on a 1060? How does that work?

But........... why? Why would you buy a Ryzen 7 CPU and disable some of it knocking you down to Ryzen 5 performance? Aftermarket coolers aren't expensive at all, and Ryzen doesn't get that hot either.

 

Btw, I have a GTX 1060 as well...

 

Anyways, my assessment is that either there's an issue with the graphics card and the CPU is trying to compensate, or there's a heat issue / throttling somewhere, bloatware / junk running in the background, bad OC / settings, or a combination of a few things which is what my bet is. Definitely not an issue with the game itself. You'll likely need to go through a process of elimination.

 

With a 12 thread CPU, you should be getting a maximum of 33% overall CPU usage with this game. The math is 400% divided by CPU threads = max overall CPU usage. 400% = 2 cores x 2 hyper-threaded x 100% = maximum the game can use. This game spreads the load to all core threads somewhat evenly to maximize stability.

 

If you're getting more than 400%, then there's other stuff running in the background.

[SylenThunder]

Not true, this game would likely not work at all on a single core. Also, AMD's version of hyperthreading works with this game now, they fixed it. So now it's basically 2x2 threads spread out on all cores for this game, just like any Intel CPU. It used to be just 2 cores with no hyperthreading, but that got fixed at some point.

I can roll with that. I know there was an issue, both with Unity and UE4. As far as I'm aware, it's still an issue with UE4.

[Fox]

I can roll with that. I know there was an issue, both with Unity and UE4. As far as I'm aware, it's still an issue with UE4.

Ya, I just tested that with Hellblade: Senua's Sacrifice (UE4) and it does seem to focus mainly on 1 core, though the other cores are still a bit active, but it's definitely not balanced. Based on my math, I was getting an average between 17% to 23% cpu usage (and one random spike to 32%), times that by 12 cores and it does go above 200%... so it does appear to use more than 1 core, but not very well. Probably the engine uses 1 core and the game devs try to use the other cores for everything else.

[LethalWarfare]

Also since I am runing multiplayer I am also hosting the server from my PC.

 

- - - Updated - - -

 

Also coolers are not that cheap. The cpu comes with a really good air cooler that works fine, But it's just not enough when overclocking it. Water cooling would work really well but that is not cheap.

[Fox]

Water cooling (in my opinion) is not worth it at all, especially those dumb all-in-one options which you can't service at all. Then again, increasing the voltage on a CPU in my opinion is also not worth it considering what little you gain in real world applications. Premium air coolers, like Noctua coolers are definitely expensive, I agree, but I think Cooler Master has some much cheaper coolers that provide almost as much cooling as Noctua and therefore plenty of room for overclocking. I can't vouch for the other name brands though which are even cheaper but look similar in design.

 

I haven't done any CPU load testing while hosting a multiplayer game so I have no information for that.

[SylenThunder]

Water cooling (in my opinion) is not worth it at all, especially those dumb all-in-one options which you can't service at all. Then again, increasing the voltage on a CPU in my opinion is also not worth it considering what little you gain in real world applications. Premium air coolers, like Noctua coolers are definitely expensive, I agree, but I think Cooler Master has some much cheaper coolers that provide almost as much cooling as Noctua and therefore plenty of room for overclocking. I can't vouch for the other name brands though which are even cheaper but look similar in design.

 

I haven't done any CPU load testing while hosting a multiplayer game so I have no information for that.

I second this. It's actually been proven that water cooling is not more effective than proper air cooling outside of some really extreme cases, and even then the tradeoff was not a large enough benefit to support it.

[Beelzybub]

I think as a test I would pull the 1050 and see what happens. Maybe there is a conflict somewhere with the two cards. I have a 1060 6gig and an i5 3550 and get 70-100 fps at 1080p. Overclocking is not where you should be looking at this point. There's something else wrong.