High resources utilisation

[caste381]

Thank you.

I have forwarded the logs to the support team via email.

[Zacchaeus]

Interestingly this weeks after the patch my player's CPU usage stayed steady at around 300%. Still high of course but considerably reduced from previous sessions.

[Zarkamorta]

Here's a benchmark for those interested (and hopefully a simple enough test many others can duplicate);

FGU launch screen, resting
~10% CPU Load total, 3.6 GB RAM utilized
FG using ~2.5%, 113 MB RAM

Creating & Opening New 5E Campaign
Peak CPU ~31%, Average ~25%, Resting ~12%
Resting FGU 290MB, ~4% CPU load

PHB, MM, DMG, DoMM loaded
Peak Total CPU ~42% CPU
Resting FG process ~5.5% CPU, 1670MB, GPU ~20%

Computer
Dell Precision M3800 (laptop)
Intel i7-4712HQ, [email protected] GHz
16 GB RAM
Windows 10 Pro, 64 bit
NVidia Quadro K1100M (Driver version 425.45)

Those are well within my performance expectations. And we can see the issues with FGU performance are not universal.

Hi LordEntrails,

What was the screen resolution for those benchmarks. I would like to replicate them on my Surface Pro 7 to compare them. I already noticed that GPU load scales with resolution. To be able to compare the results, I would like to run the benchmarks at the same resolution. And could you maybe add a benchmark with a map open, because this puts extra stress on my GPU. Further, do you have FPS counts and GPU load for those benchmarks? Though, since your GPU stays at max 20 % I would guess it should be 60 FPS?!

In theory, my GPU should be faster than yours, but this is definitely not the case in FGU. I have massive problems running the software.

https://gpu.userbenchmark.com/Compar...10244vsm921079

[LordEntrails]

Hi LordEntrails,

What was the screen resolution for those benchmarks. I would like to replicate them on my Surface Pro 7 to compare them. I already noticed that GPU load scales with resolution. To be able to compare the results, I would like to run the benchmarks at the same resolution. And could you maybe add a benchmark with a map open, because this puts extra stress on my GPU. Further, do you have FPS counts and GPU load for those benchmarks? Though, since your GPU stays at max 20 % I would guess it should be 60 FPS?!

In theory, my GPU should be faster than yours, but this is definitely not the case in FGU. I have massive problems running the software.

https://gpu.userbenchmark.com/Compar...10244vsm921079

1920 x 1080

Don't have frame counts etc. I'll see if I can redo this and try and capture more numbers. Also with opening some maps.

If you take this test and extend it with steps, I'll reproduce that. Whoever gets there first

[Zarkamorta]

Here's a benchmark for those interested (and hopefully a simple enough test many others can duplicate);

FGU launch screen, resting
~10% CPU Load total, 3.6 GB RAM utilized
FG using ~2.5%, 113 MB RAM

Creating & Opening New 5E Campaign
Peak CPU ~31%, Average ~25%, Resting ~12%
Resting FGU 290MB, ~4% CPU load

PHB, MM, DMG, DoMM loaded
Peak Total CPU ~42% CPU
Resting FG process ~5.5% CPU, 1670MB, GPU ~20%

Computer
Dell Precision M3800 (laptop)
Intel i7-4712HQ, [email protected] GHz
16 GB RAM
Windows 10 Pro, 64 bit
NVidia Quadro K1100M (Driver version 425.45)

Those are well within my performance expectations. And we can see the issues with FGU performance are not universal.

So here are my benchmarks in comparison on a Surface Pro 7 (which GPU is more powerful than the Nvidia Quadro K1100M, all benchmarks are at the same resolution of 1920x1080):

Win 10, resting
CPU, 1%, 3.4 GHz, 2.8 GB
GPU 0%, 0.1 GB

FGU launch screen, resting
CPU 9%, 3.5 GHz, 2.9 GB
GPU 25%, 0.3 GB, 60 FPS

FGU New 5E Campaign, No Extensions, LAN, scaleui 100, resting
CPU 13%, 3.5 GHz, 2.9 GB
GPU 32%, 0.2 GB, 60 FPS

FGU New 5E Campaign, as above+PHB, MM, DMG, FG Battle Maps, resting
CPU 13%, 3.5 GHz, 3.3 GB
GPU 33%, 0.2 GB, 60 FPS

FGU, New 5E Campaign, as above, BattleMap_Cave01 open, resting
CPU 10%, 1.0 GHz, 3.6 GB
GPU 98%, 0.3 GB, 42 FPS

So basically my GPU already sees a higher load in the 5E campaign, though it should be more powerful than yours. The moment I open a map the GPU jumps to 100% and starts throttling the CPU and the FPS decrease.

I still think there is something faulty with how FGU handles modern GPUs, no matter if they are integrated or dedicated. I can play Fortnite on my Surface Pro 7 at 45 - 60 FPS and this should be more demanding than FGU. Also a load of 15 % on a GTX 1660 Ti or 50 % on a GTX 1050 Ti (other post) sounds rather high compared to what FGU is graphically doing.

I still hope that the Dev will look into this and fix it.

[Tuleen Donai]

Interestingly this weeks after the patch my player's CPU usage stayed steady at around 300%. Still high of course but considerably reduced from previous sessions.

Did you really mean 300%? :-) And, that's reduced?

[LordEntrails]

Did you really mean 300%? :-) And, that's reduced?

yep, people were reporting over 700%

[caste381]

I still think there is something faulty with how FGU handles modern GPUs, no matter if they are integrated or dedicated. I can play Fortnite on my Surface Pro 7 at 45 - 60 FPS and this should be more demanding than FGU. Also a load of 15 % on a GTX 1660 Ti or 50 % on a GTX 1050 Ti (other post) sounds rather high compared to what FGU is graphically doing.

That's exactly my point. As good as FGU is, it is a "simple" piece of software, graphically speaking. Such a high use of resource is very difficult to justify.
As said in another post, I do not see why an empty campaign with no users connected should use more than 1-2% of CPU.

[LordEntrails]

That's exactly my point. As good as FGU is, it is a "simple" piece of software, graphically speaking. Such a high use of resource is very difficult to justify.
As said in another post, I do not see why an empty campaign with no users connected should use more than 1-2% of CPU.

Resource utilization is not justified, and I don't recall anyone ever trying to. The devs have stated time and again they are working on performance. They have always stated, and users here have proven, that performance is not universal to all platforms.

The devs are and will keep working on performance tuning. If you want to help, provide detailed system specs, usage cases and all the other info so that they can duplicate your issue (if possible) and so that they can figure out what the difference are between systems that see better performance and worse performance. The need help (data) to identify what programs, dlls, hardware, settings or other differences that account for the difference performance seen by different users.

[caste381]

Resource utilization is not justified, and I don't recall anyone ever trying to. The devs have stated time and again they are working on performance. They have always stated, and users here have proven, that performance is not universal to all platforms.

The devs are and will keep working on performance tuning. If you want to help, provide detailed system specs, usage cases and all the other info so that they can duplicate your issue (if possible) and so that they can figure out what the difference are between systems that see better performance and worse performance. The need help (data) to identify what programs, dlls, hardware, settings or other differences that account for the difference performance seen by different users.

I have provided info in the previous posts and sent system logs and specs via email to support.
Good to know you also believe high utilisation is not justified.
I have seen some post talking about upgrading hardware...