Battlefield V only uses 80% of the cpu and does not reach max turbo frequency

I recently got an MSI GE63 8SG, with an i7-8750H cpu and RTX 2080 gpu.

The problem I'm having is when playing Battlefield V:

• In the main menu, the cpu boosts to 3.9 GHz - the max turbo frequency - as expected.
• But during gameplay, the cpu sits on constant 3.0 GHz and 80% usage, never going past these values.

The GPU is nowhere near 100% usage either, so it's not like the cpu is being bottlenecked by it.

I already undervolted the cpu and have good temperatures while playing the game - 70~75ºC normally, and 60~65ºC if using cooler boost.

I've been monitoring the hardware using HWiNFO and the cpu is not being throttled due to temperature / power / etc.

I've also stress tested the cpu when undervolting and it behaved well, performing at max turbo frequency (3.9 GHz) while never going above 80º C.

How can I make the game boost the cpu to its max turbo frequency and use its full potential? This laptop should be able to get much higher fps than what I'm currently getting, which is barely 60 and not at all constant.

 

What is your RAM?

 

Use GPUz to see if the GPU is power throttling. That is my next guess.

I have a Cyberpower Tracer III laptop in my posession whose GPU power throttles unless I set it to it's performance mode within Cyberpower's gaming center application, which is essentially their version of a control center. Does MSI have something like that? If so I'd take a look at your settings.

Also, Battlefield 5 gets significantly better framerates with RAM in dual channel mode. If you only have one stick, consider equipping your laptop with another.

 

I have 16Gb of RAM. The GPU is not throttling.

I don't think this is a RAM problem - when I tried the stress tests the cpu performed perfectly.

It doesn't matter if I'm playing on low or ultra settings in the game, I always have the same fps and cpu usage. The cpu is always at 80% and nowhere near its maximum clock speed.

 

In fact, while MSI Afterburner tells me the cpu is at 2.9~3.0Ghz, HWiNFO's Average Effective Clock displays 2.3~2.5 GHz, which is even crazier when I should be actually reaching ~3.9 GHz

 

Look at this benchmark of my exact laptop playing the game on Ultra settings:

That person's cpu (again, exactly the same laptop as me) is boosting at 3.9 GHz as expected, but mine does not.

 

Is the RAM single or dual channel?

 

If you are not sure, CPU-Z will tell you.
Single-channel is going to suck in more things that BF5. There's really no excuse for running any computer in single-channel mode, but there are a lot of turdbooks that are being sold that way.

 

It's single channel 16 Gb. Although I agree that dual channel is better, is it really that likely that the reason 20% of the cpu is not being used is because of single channel RAM? The cpu does boost properly to max turbo frequency during stress tests and during the game menus - just during actual gameplay it sits at 3.0 GHz (2.5 according to hwinfo), independently from low or ultra graphics.

 

Single vs. dual channel makes a massive difference in Battlefield V.

 

Yes. You'd be better off with 8GB of RAM in dual-channel than 16GB or 32GB in single-channel. Your computer is going to have lackluster gaming results in single-channel mode until you fix that. It never should have been sold to you like that. It is totally absurd that any company is selling notebooks with single-channel memory configurations, but it seems to be a popular trend.

 

Look in the picture you posted how the cpu usage and clock speeds are identical in both cases. I understand I will get better fps with dual channel RAM - that's not the point. The point is that the cpu is barely boosting and not reaching the clock speeds it should.

 

Because your CPU is being severely bottlenecked by its lack of memory bandwidth in single channel.

 

I see. I will give it a try and see if it improves. Thanks for the help!

 

Watch this video to see how a PC with otherwise decent specs can be bottlenecked by single-channel memory.

 

The difference is critical in edge cases with small amount of RAM, e.g. 8GB - although 16GB is not that much either, nowadays. With enough RAM the difference is nowhere near dramatic.



Bottom line, I very much doubt single channel 16GB RAM stick is the root of @tirafesi 's problem. While adding another stick will improve gaming performance, it is unlikely to solve the described issue.

 

I think the CPU utilization is merely assumed to be an issue. The underlying problem causing speculation about that is low FPS. Single-channel is a problem and needs to be fixed if performance matters. Once that global problem is eliminated we can see if low FPS continues to be a sticking point. Always best to approach problem-solving methodically and single-channel memory castration is the best and easiest place to start. That needs to be fixed even if it doesn't fix the BF5 problem.

 

Ok, I found the problem! 'Windows performance power' (I didn't even know this slider existed - first time having windows 10 on a laptop...) was set to 'better performance'. As soon as I increased the slider to 'best performance' the cpu started boosting to the max turbo frequency.

 

The Frostbite Engine loves memory bandwidth. It even benefits from quad channel memory:



Most of the games tested in those videos were in GPU-bound scenarios. When the GPU is the bottleneck, the CPU/RAM makes no difference.

 

A bit off-topic, but it really interesting that dual channel has a big performance boost over single channel, though quad channel has little to no impact over dual channel

 

You should have started saying that, lol

 

That is a subjective comment. It really depends on what you are doing. Quad channel can make a huge difference, but usually not in gaming. Same as high core count and massive overclock don't usually do a lot for gaming.

 

lol, yeah I was talking in gaming performance. I'm always talking about gaming performance lol.... About the core count and overclocking. As long as game engines uses fewer cores the core count and overclocking should have little effect. Most modern CPUs already do a large overclock when few cores are being utilized. Hopefully now that consoles have 8 cores CPUs we will have a better CPU utilization.

 

When games user fewer cores, overclocking matters because it's about the per-core performance (frequency and IPC).

 

Yeah but turbo boost does that already. The CPU automatically overclocks based on the core count usage activity. That's why sometimes for gaming leaving the CPU at stock gets better fps than overclocking all cores. You can of course also overclock all your cores past turbo speed or you overclock the turbo boost table. In general I'd say that turbo boost is ok for gaming.

A good example would be a CPU with turbo between 4.0 - 5.2 GHz, if the user overclock all cores to let's say at 4.7 GHz that user might get worst performance than leaving the CPU at stock.

 

This might apply to Ryzen since their boost algorithm works differently, but not to Intel CPUs. On Intel CPUs, especially mobile, the single core boost is often much higher than the all-core boost. Even in lightly threaded games, the CPU will spend most of its time at or near the lower all-core boost since the Windows scheduler will assign background tasks to the ancillary threads.