It is sad, I don't play it much because of performance issues which seems to directly affect hit detection.
PlayerUnknown's Battlegrounds (PUBG) CPU Test
- Thread starter Steve
- Start date
It is sad, I don't play it much because of performance issues which seems to directly affect hit detection.
I tried 2133mhz ram for a while, and moved to 2666mhz and now its at 2800mhz OC and it didn;t seem to make a huge difference in performance on my 5820k at 4.7ghz, Though Cache frequency seemed to make a bit of difference, although it could of been a better server, its hard to test as its so different everytime.
cliffordcooley
Posts: 13,141 +6,441
Did you read the article. The author said the same thing. Perhaps you could ask him. It was somewhere alone the lines of all other games getting at least twice the frame rate at the same resolutions. And lets not forget the differences in Intel and Ryzen. This game is clearly not optimized for Ryzen on top of the bad performance with Intel.
GreenNova343
Posts: 454 +336
Maybe because in order to hit that mark the Pentium needs the GTX 1080TI, one of (if not the) fastest GPU currently available to gamers, & a GPU designed for very high framerates @1440p/60FPS @ 4K gamplay? Or because its big brother i5-7600K (Kaby Lake, so same generation; only slightly faster clock speeds, & while it has double the cores it has the same number of threads) is able to get 60% more average/80% more minimum performance with the same GPU?
Given the choice, would you spend a little over $1,100 USD for a Pentium G4560/GTX 1080TI build (https://pcpartpicker.com/list/YzLbFd), $100 less for an i5-7600K/GTX 1070TI build (https://pcpartpicker.com/list/g9dJsJ), or even $30 less to pair the 1070TI with an i7-7700K (https://pcpartpicker.com/list/46bVsJ)? Or even better, spend just under $1,000 USD for an i5-8400/GTX 1070TI build (https://pcpartpicker.com/list/xGFYVY)? Which do you think is going to be the better overall gaming value?
GreenNova343
Posts: 454 +336
One "loss" in a badly-optimized, still-in-extended-beta-testing game is very far from calling Ryzen a flop.
justhateme
Posts: 6 +0
-useallavailablecores - works. Those ones above (threads and cpucount) not in registered launch options for pubg
Vulcanproject
Posts: 1,701 +3,212
Ryzen flops hard in this game. Pretty undeniable looking at the tests as my next sentence pointed out. Context my good fellow, get some.
It's not the only game Ryzen loses quite badly on, but it's the one tested here and also just so happens to be the most popular PC game on the planet right now by some distance. Not insignificant.
cliffordcooley
Posts: 13,141 +6,441
So you can attest those games were optimized for Ryzen? Because then that would be the only clear evidence Ryzen sux.
Abraka
Posts: 176 +54
Are you blind? Or just ignorant? It's not AMD's fault that greedy fools are deliberately not using the available power. AMD CPUs are basically on vacation while playing this game. With Ryzen CPUs you can mine Ethereum while playing this game on ultra settings and you won't notice any drop in performance.
It's clear that this particular stinking game uses only 3 threads. Not 4, as someone said, but only 3. Knowing the history of Intel, and their recent scandals, I wouldn't be surprised if they paid the game maker to limit the thread usage. Because that limitation hits AMD CPUs much more, since entire previous line of Intel CPUs had a maximum of 4 cores. That's why 7700K fares so well.
I'd say this is one of many clear cases of corruption in the gaming industry, where game makers get sponsorship from a CPU or GPU maker, sometimes even openly advertising it (remember the Nvidia logo at the start of some games?)
As a player, I hate that, because that means it may under-utilize my hardware, depending whether the sponsor is Intel, AMD or Nvidia. If I bought i5-8400 or Ryzen 1600, most of my cores will sleep unused. Well, I'm not paying to greedy bastards. I'll rather buy another tank in World of Tanks than support botch like PUBG.
Not only they are greedy, but look at the graphics quality. 10 years older games had better graphics than this. And run faster too.
Vulcanproject
Posts: 1,701 +3,212
I can read the graphs is what.
Whether or not they are optimized for Ryzen was besides my point. My point is AMD sucks on this game compared to Intel and it's a big, big game. It matters a lot to gamers it seems.
However if we go into it people raging about that fact and AMD fans wailing and beating their chests does not help the situation. I read all this over and over and over the past decade since Piledriver and it's just as useless an argument today. It's always a conspiracy, and it's always everyone else's fault for not spending forever optimizing and building their game specifically for whatever AMD come up with. The actual question you have to ask is why would they? They are a tiny minority of the gaming market. AMD is 8 percent according to Steam survey. Presumably Ryzen as an arch is an even smaller fraction of that.
For everyone playing this game and there are millions at a time daily, that talk doesn't actually do anything for them though does it? It's the evil Intel illuminati, it's those evil developers targeting by far and away the largest segment of the market etc.
Maybe it's just that AMD's chips aren't as good for gaming, and maybe they should look into it and work on that hard internally, and in cooperation with developers that make the most popular games. Then we can sit here and congratulate their success on matching Intel's gaming performance without invoking the same smoke and mirror excuses I have read for years now.
F
Flebbert
Lets leave the sarcasm at home.
The patch due out on the 12th of December is referred to as patch 1.0 which everyone who plays PUBG is aware of. I swear this place is more about measuring members than the tech these days.
TLDR: Game still poorly optimized, not a pissing contest between Intel and AMD.......... Everyone who plays PUBG just wants it to run better for everyone, Oh and less cheaters in the game would be nice.
cliffordcooley
Posts: 13,141 +6,441
OK Thanks for clearing that up. I will apologize for my inconsiderate response.
Abraka
Posts: 176 +54
Obviously you know nothing about computers. It's not on AMD to optimize their CPU for every freaking game that appears (and let's face it, this game sucks in graphics). It's up to good programmers to utilize the power of the machines they are running on. If they can't use multiple cores it's the fault of stupid programmers. AMD gave them ton of cores. Intel gave them 6 cores in the latest update. And they use only 3 in the game. So, obviously, game is crap.
Now, older versions of AMD CPUs weren't even close to Intel, so with or without optimizations there was no way for AMD to compete. But with the current generation it's clear that Ryzen 1700 is faster than i7-7700K. It's got more raw computing power. And yet, the game can't use it.
Who's fault is it? Certainly not AMD's. If the CPU load is only 28% it's clear that CPU cores aren't utilized. Now, if the game makers can't find good programmers who can write multi-threading code, maybe it's time to offer honest salary, so good programmers would apply. It's not the AMD's fault that game owners are cheap bastards.
I am running a 4690K with a GTX1080 and I'm struggling to hold 60FPS on Ultra....
justhateme
Posts: 6 +0
The results are quite strange. When I was running "Pochinki test" on 7700k 4.6GHz, 1080ti and 16Gb ram it was like 160 top fps, 115 top 20%, and everything else below 80. During loot spawn it was even lower then 50.
It's 100% horribly bad optimisation. You can't just look at one result, you have to look at the overall numbers. And the game isn't getting "over 60FPS", it is just an average. You'll regularly drop to 40-50 FPS in more busy scenes when you aren't just walking or looting.
If playing a shooter type game with the lowest settings at 1080p, with the best gaming PC money can buy, you can't get past 120FPS then you can objectively call it crap in terms of optimisations. Especially when it looks like it is a Xbox360/PS3 title.
Last edited:
gamerk2
Posts: 983 +996
GPU bottleneck. Only so much you can do to make the CPU go faster if the GPU is hitting a wall at ~122 FPS or so.
As for Ryzen, PUBG isn't doing that much CPU side that could be threaded to benefit the architecture. You don't have high end graphics, AI, dynamic weather effects (a rain filter doesn't count), or a robust physics engine. There really isn't much for the CPU to do, so it follows that clock * IPC drives CPU performance.
Anton Skryaga
Posts: 38 +18
I liked that, it's true.
But another points is here:
1. Intel CPUs can run higher freqs today.
2. Not every task is possible (or simple enough) to rewrite to be (efficiently) multi-threaded.
So Intel's CPUs are better for modern desktop software (IMO).
AMD is more like on strategic way to future - we will use many cores anyway.
(Yes, I'm Intel/Nvidia biased and never before used AMD for my systems)
Vulcanproject
Posts: 1,701 +3,212
Obviously you missed the point. You also seemingly fail to realize that this game runs on a licensed engine. Maybe you don't know how this works quite that well. It isn't clear that 1700 is faster than 7700K at all. 'Raw computing' theoretically, another word for meaningless in the real world unless deployed.
It's up to AMD to build a CPU that works well in all the functions it is intended to perform. If they fail in that aspect (may I direct you to weak Ryzen integer performance and cache hierarchy problems allied to a host of gaming benchmarks) then unless it's a massive share of the market a developer does not have a large incentive to work specifically well with a particular CPU arch. No doubt they use AMD's developer kits, but it's not up to them to try and 'fix' those to work better either. It's up to AMD to work with them.
Epic look at average machine specs and build the engine to suit, the developer builds the game with the engine. The developer relies on the licensed engine and that engine's performance on a given arch. Now ask yourself, if AMD's CPU does not perform very well on particular engines compared to it's rival because of architectural deficiencies and it makes up a small volume of the market, largely BECAUSE of said deficiencies, who should have the MOST incentive to improve that in games?
The developer? Not really. Tiny percentage of their end users. The engineers that create the game engines? Partially, but they target a market segment by looking at hardware penetration. AMD?
Absolutely. Because if they WANT their processors to sell better and break the chain where their rival's hardware is preferred, they are the ones that have to spend the money to work with engineers and developers to improve performance.
So instead of sitting around moaning about it there is plenty AMD can do about it which apparently they aren't doing. Both inside and support outside the company.
Last edited:
Pastuch
Posts: 155 +169
Wow! This game runs like dog sh!t on every piece of hardware out there. I have an 4.4ghz Intel chip with a 1080ti and I still cant pass 120 fps at 1080p which is atrocious. On my primary 4k monitor (55 inch lg oled) it lags brutally, regularly into 20fps! Compare that to Shadow of War which looks astonishingly beautiful in HDR 4k and NEVER drops below 60fps. The optimization is f'ing terrible. Oh and CPUs dont matter at 'adult' resolutions. Buy a decent display!
Pastuch
Posts: 155 +169
Option 4: keep using my 4.4ghz 2500k. Buy a cheap 1080ti and then spend all my money on a 55 inch 4k oled with epic HDR color. They are down to $1400 for a 55b7A which completely embarrasses ever computer monitor I've ever had. Cpus are useless at 4k. Get a 1080ti and an LG OLEd.
Mustafa OZCAN
Posts: 9 +4
Another elaborately article by Mr. Walton..
I always enjoy reading your writings. It's clear to see the efforts all the time.
Have a nice day..
a TechSpot / Hardware Unboxed fan from Kastamonu, TURKIYE
(sorry for bad english)
I always enjoy reading your writings. It's clear to see the efforts all the time.
Have a nice day..
a TechSpot / Hardware Unboxed fan from Kastamonu, TURKIYE
(sorry for bad english)
Were these tests done with the 144hz default cap disabled? It is set per default and has to be set higher via start parameter.
Seeing that framerates for very low, medium and ultra are almost the same, I'm afraid your GPU hit 144 FPS a lot with some CPU or RAM (yes, PUBG scales well with RAM) related drops, pulling down the average a bit.
Seeing that framerates for very low, medium and ultra are almost the same, I'm afraid your GPU hit 144 FPS a lot with some CPU or RAM (yes, PUBG scales well with RAM) related drops, pulling down the average a bit.
