[SOLVED] What's causing the bottleneck?

[sampaudel0]

My System:
Processor: i5 4690
GPU: MSI RX 570 8GB
RAM: HyperX Fury 1x8 GB DDR3 1866 MHz
Motherboard: MSI B85-G41

Ever since I started playing e-sports titles after upgrading my graphics card from a GT740, my frames have stayed relatively low. I know it's outdated and I thought the issue was the CPU, but after looking online on bottleneck calculators (~2%) and forums, performance seemed much better than what I get.
E.g., Fortnite gets ~80fps on lowest settings, Overwatch also stays around 80 on low. Even Minecraft seems oddly low at around 100fps. Since I've saved enough to replace my CPU, Mobo, and RAM, I'm starting to think that the RAM speeds may also be the issue. Is this the case? Thanks

 

[Phaaze88]

From just looking at those specs, the single stick of ram raises a flag, although there could be other things happening in the background while in-game.
A single stick of ram hurts your fps by way of lower minimums and averages.

Home - UserBenchmark

Go here^and POST a LINK to your results.

 

[sampaudel0]

From just looking at those specs, the single stick of ram raises a flag, although there could be other things happening in the background while in-game.
A single stick of ram hurts your fps by way of lower minimums and averages.

Home - UserBenchmark

Go here^and POST a LINK to your results.

That makes sense. So do you think upgrading my parts to something like: Ryzen 5 2600 and 2x8 GB DDR4 3000MHz wield similar results as just upgrading my DDR3 RAM to 2x8 GB?

 

[Phaaze88]

Do your games max out all four cores? If so, the extra cores/threads of the Ryzen 2600 + 2x 8GB 3000 ram, would yield larger gains over just 2x 8GB DDR3s.

Do your games just continue to peg one or 2 cores at 100%? If so, the Ryzen 2600 combo may not offer that much more of an improvement over the 2x DDR3s, unless you get Ryzen 3600 instead.

If you experience both, I'd say the 2600 combo would be better overall.

 

[sampaudel0]

I have my RX570 sent in for repair at the moment so with a GT710 (I know, massive bottleneck), I get these values:

CPU #0: 62%
CPU #1: 57%
CPU #2: 59%
CPU #3: 46%

Average FPS: 40-50
Resolution: 480x270
GPU Usage: 100%

So it seems like they all are being used, guess the 2600 upgrade would be the best right?

 

[Phaaze88]

Is your monitor's resolution only 480x270?
Then I have to take back what I posted earlier.
Aside from the lack of a second stick, a cpu upgrade isn't going to change much here.

A PC is only as good as it's weakest link, and in this case, it's actually the monitor.
As you go up in resolution, gpu load increases and cpu load decreases. Vice versa, it's the other way around.

You should check your results with the RX 570 when you get it back. It is much stronger than a GT 710, and I imagine it's not doing a whole lot at that resolution.
You can put a cpu upgrade on hold - a higher resolution monitor would be more practical.

 

[Karadjgne]

Eh.
Start from scratch. Very simple test. Play the game for 5 or so minutes at current settings and log the average fps. Restart the game with ultra high max settings. Log the average fps.

The cpu sets the maximum fps limit. It pre-renders all the frames before shipping them to the gpu. The gpu either can do those frames or not.

So, if fps barely changes, if at all, then the gpu is strong enough for the job and the cpu is maxed out limit of fps. If the fps tanks hard, then the cpu is fine, it's the gpu that cannot reproduce everything its given.

Common issues with cpu are overheating, lack of ram, storage, usage and motherboard chipset drivers.

Common issues with gpu are overheating, lack of vram, drivers.

So first figure out which of the 2 is the issue, cpu or gpu. Then back-track from there to figure out what is the likely root cause creating that issue.

This can mean using DDU to fully fix gpu drivers, going to motherboard website and fully upgrading any/all chipset, audio, Lan, USB family, Sata, pcie, Intel management engine drivers, doing a full anti-virus and malware scan (they are different), using ccleaner (piriform.com) default settings to do several cleans and the registry tool as well (default, say Yes to backups!), not to mention using realtemp/coretemp/Afterburner to monitor cpu/gpu temps.

Single or dual channel ram won't make any meaningful difference except for one thing. IF the bandwidth of the single channel is saturated (doubtful) it'll run @ upto 20% slower than dual channel. Otherwise there's no difference.

One other thing to consider is that Fortnite is server bound. Your cpu is at the mercy of every other player in your game. The more players, the harder it is on the cpu to maintain fps in certain areas with multiple foreground movable objects. Fly text, names, objects etc all cpu killers. I play SWtOR online with a 3770k/970. Single or 8man team, no issues. 16man and I need to turn off fly text and bloom. 24man world boss fights are miserable choppy, just too much happening with special affects graphics and physics from all the different attacks and results. Fps goes from beyond 60+ to 30 ish.

 

[sampaudel0]

Is your monitor's resolution only 480x270?
Then I have to take back what I posted earlier.
Aside from the lack of a second stick, a cpu upgrade isn't going to change much here.

A PC is only as good as it's weakest link, and in this case, it's actually the monitor.
As you go up in resolution, gpu load increases and cpu load decreases. Vice versa, it's the other way around.

You should check your results with the RX 570 when you get it back. It is much stronger than a GT 710, and I imagine it's not doing a whole lot at that resolution.
You can put a cpu upgrade on hold - a higher resolution monitor would be more practical.

Actually not the case. When I tried playing on the GT710, I could only really get 60fps on that res, I have a 1080p monitor. I can post the results with 1080p tomorrow morning. I’d expect something like 10-20 FPS.