PlayerUnknown's Battlegrounds (PUBG) CPU Test

[Julio Franco]

Following up on the mini-test we did for PlayerUnknown's Battlegrounds back in June, it was about time we checked where the game's performance is at after receiving countless updates. Focusing on CPU performance, we have all 8th-gen Core processors, all Ryzen CPUs, and a few from the 7th-gen Core series.

Read the full article here.

https://www.techspot.com/article/1532-pubg-cpu-benchmarks/

 

[treetops]

I wonder where the 4690k would rank on this.

 

[Puiu]

TL;DR the results are pathetic for a game this popular.

We already knew that this game is incredibly bad when it comes to optimisation, but I was at least expecting a few good changes after so many patches. This is just sad.

 

[Flebbert]

I cannot believe how sh!t this is and really quite disappointing with how much money they have and people working on this.

I will be really disappointed if the patch 1 release does not resolve the performance issues.

Also kind of annoying that they are not even close to supporting multi-core and multi thread CPU.......... Going to call that one as hog wash.

I am a little frustrated with the developers

 

[cliffordcooley]

I will be really disappointed if the patch 1 release does not resolve the performance issues.

Lets go back and read @Puiu comment again. You know the one that you liked.

We already knew that this game is incredibly bad when it comes to optimisation, but I was at least expecting a few good changes after so many patches. This is just sad.

Puiu makes the assertion there has been many patches.

TLDR: So Flebbert, you don't have to wait to be disappointed. Patch 1 is history.

 

[R00sT3R]

I wonder where the 4690k would rank on this.

Just knock 5-8% off the 7600K results, perhaps 10% if you are running slow 1600Mhz DDR3 with a 4690K compared to someone running fast 3200Mhz DDR4 with a 7600K, and there is your answer.

In other words, hardly any difference that you'll notice and still perfectly fine for this game.

 

[Anton Skryaga]

Looks like PUBG can start up to 4 main execution threads.

Let's check several calculations:
Ryzen 7 1800x: 4 of 16T = 25% this is max possible load for this CPU (24% in test) .
Ryzen 5 1600: 4 of 12T = 30% (30% in test)
Ryzen 5 1400: 4 of 8T = 50% (40% in test)
I think we see here like Ryzen's four cores running PUBG hit 100% load each.

i5 8400: 4 of 6T = 75% (50% in test)
i5 7600: 4 of 4T = 100% (72% in test)
And here Intel chips reached GPU's cap.

This version can be checked by testing same CPUs with HT disabled and active cores reduced to 4 (don't know if it possible on desktop mobos) - results should be close to those we see in this article.

 

[Banan]

It's clearly visible that frequency is more important than core count in this game. I would like to see overclocked Pentium G3258. It will probably crush the new expensive CPUs

 

[BSim500]

I don't own the game but it's great to see Techspot test different presets beyond "Ultra-only" of many other sites. I hope this continues to be a trend for future individual game performance analysis articles.

 

[dj2017]

It probably uses 4 cores, maybe partially using 2-4 more cores for little stuff like extra filters and effects when increasing visual quality in settings. it's more obvious when comparing the 7600K and 8400 results.

Power consumption numbers would be interesting here.

 

[max0x7ba]

It would be interesting to see the benchmark results with AMD RX Vega 64 LC, its driver utilizes the CPU differently.

Also, there are reports that the game benefits from fast RAM, it would be interesting to see results with DDR4 4000MHz.

 

[Strawman]

Nice testing, well done sir.

The most CPU intensive area of the game is Yasnaya Polyanna btw. In that area, the whole Ryzen series struggles (and I kid you not), to sustain mid 40 fps, even in ultra low quality. Even an overclocked to 4.6ghz 4790k barely hits 50. I would be interested to see how the coffeelakes do in that. I doubt they can hold 60 either.

 

[dirtyferret]

TL;DR the results are pathetic for a game this popular.

We already knew that this game is incredibly bad when it comes to optimisation, but I was at least expecting a few good changes after so many patches. This is just sad.

Why because AMD lagged behind Intel as usual? I see pretty good fps regardless of CPU but also a game dependent on IPC like virtually all other games out there.

 

[Anton Skryaga]

Quick googling shows PUBG have an options for threads/cores: -cpuCount=4 -threads=4
and -USEALLAVAILABLECORES (unconfirmed)

Can somebody check this info?

 

[Puiu]

Why because AMD lagged behind Intel as usual? I see pretty good fps regardless of CPU but also a game dependent on IPC like virtually all other games out there.

Pretty sure I didn't mention AMD or any company. If you hate AMD or having a fanboy moment then that is your own problem.

Are you telling me that the best gaming PC money can buy cannot get a steady 120FPS at 1080p with the lowest settings in a game that looks like it belongs on a last gen console? Sorry dude but your "pretty good FPS" is not objectively true.

What is the point of having a 144Hz monitor then if not for this type of games? (not to mention 1440p 144Hz monitors)

 

[dirtyferret]

Pretty sure I didn't mention AMD or any company. If you hate AMD or having a fanboy moment then that is your own problem.

Are you telling me that the best gaming PC money can buy cannot get a steady 120FPS at 1080p with the lowest settings in a game that looks like it belongs on a last gen console? Sorry dude but your "pretty good FPS" is not objectively true.

What is the point of having a 144Hz monitor then if not for this type of games? (not to mention 1440p 144Hz monitors)

Your MO on this forum has been to piss and moan about anything that doesn't put AMD in a good light. You are not fooling anyone!

 

[Kevin82485]

It seems crazy, but PUBG is technically still in Beta, so these optimization issues aren't really that surprising.

 

[Puiu]

Your MO on this forum has been to piss and moan about anything that doesn't put AMD in a good light. You are not fooling anyone!

I was always fair and impartial. I've always quoted benchmarks and did the numbers. I always gave Intel credit where credit was due.

The fact that you can't get away from the Intel vs AMD argument even now says a lot about you. If you can't stay on topic then at least don't insult ppl. Now please return to the topic at hand. You haven't even answered my 2 simple questions.

 

[Vulcanproject]

Ryzen flops. An 1800X doesn't match a 7600K.

I hope AMD can tweak Ryzen over the next year to truly offer Intel matching performance in all applications.

 

[Puiu]

It seems crazy, but PUBG is technically still in Beta, so these optimization issues aren't really that surprising.

Version 1.0 should release this December if it doesn't get delayed again. The game also has another huge problem... its net code is one of the worst (if not the worst) seen in such a big multiplayer title. It has a dynamic tick rate that can go as low as 8. Even with good ping to the server you still die after taking cover because of how incredibly slow the server gets updated.

here's a video that explains with more details:

 

[dirtyferret]

I was always fair and impartial. I've always quoted benchmarks and did the numbers. I always gave Intel credit where credit was due.

The fact that you can't get away from the Intel vs AMD argument even now says a lot about you. If you can't stay on topic then at least don't insult ppl. Now please return to the topic at hand. You haven't even answered my 2 simple questions.

Lol now that first line was funny!

For the rest of your nonsense, you are the second biggest AMD fanboy on the forum behind hard reset.

 

[gamerk2]

TL;DR the results are pathetic for a game this popular.

We already knew that this game is incredibly bad when it comes to optimisation, but I was at least expecting a few good changes after so many patches. This is just sad.

How is "getting over 60 FPS on Ultra with a Pentium class CPU" considered bad optimization? Or would you you prefer 100% utilization and 30 FPS instead?

 

[gamerk2]

I cannot believe how sh!t this is and really quite disappointing with how much money they have and people working on this.

I will be really disappointed if the patch 1 release does not resolve the performance issues.

Also kind of annoying that they are not even close to supporting multi-core and multi thread CPU.......... Going to call that one as hog wash.

I am a little frustrated with the developers

Ultra settings for the dual core Pentium are over 60 FPS on Ultra. How is that poor optimization? Or would you rather all cores at 80%, 30 FPS as an example of "good" optimization "because it's threaded well"?

Seriously, stop it.

 

[Capaill]

It's clearly visible that frequency is more important than core count in this game. I would like to see overclocked Pentium G3258. It will probably crush the new expensive CPUs

It's clearly visible that frequency is more important than core count in this game. I would like to see overclocked Pentium G3258. It will probably crush the new expensive CPUs

Hmm, 2 identical posts from 2 almost unused accounts. Is TechSpot using bots now? Surely we generate enough back-and-forth arguments without needing to add fake comments.

 

[Stardude82]

Would have like to have seen a little bit on RAM speed. You did something like this for Fallout 4 (mostly by accident).

(The G3285 would suck, obviously. Too bad Techspot didn't look at this also. Seems it work up to 6 threads.)