Steam Survey Update: It's All About Quad-cores, NVIDIA and Windows 10

[Vayra86]

PUBG is a demanding game, you need to have a really good PC to run it properly, I agree that the other 2 are not demanding but that does not mean that automatically most of the players that are playing those games have low to mid end hardware

Well... if there is one conclusion to draw from the Steam Survey, it is exactly that due to the sys requirements of the most played games in the Steam libraries, there is no need to have a faster system than mid range.

All the Steam Survey shows is that the most common denominator of performance level is what the most popular games are coded for, and this is nothing else but common sense, because that's where the greatest market potential is for a publisher.

Retrospective indeed. Steam Survey is a great way to see stagnation, you will never see progress here. In the same vein, this is also the reason we're still using quad cores for gaming; there is no argument to have more than that because games don't use it, and because nobody has any need to get more than 4 cores, there will not be many games that use it in the future either.

Another important omission from the Steam Survey is that big publishers have their own platform and nobody really likes using two layers of DRM. So Steam misses out on a vital part of the market, most notably the part where you are rewarded for having a faster system because the games can make use of it.

 

[Prima.Vera]

*Looks at system specs*
*Looks at sig*

Oh crap, i'm NORMAL

Dude, we almost have the same specs.
Almost.
But you have to tell me what black screens are you referring to the 1080 gtx and what are your freq now?

 

[Kyuuba]

There are plenty of users with 1080tis and yet still doesn't show, weird...

 

[newtekie1]

Well... if there is one conclusion to draw from the Steam Survey, it is exactly that due to the sys requirements of the most played games in the Steam libraries, there is no need to have a faster system than mid range.

Really, to play pretty much any game out, you only need a mid-range system.

The most popular display resolution, by far, is 1080p at 48%. And 1366x768(AKA laptop users) is the next most popular at 21%. But 1440p is barely over 2% of users, and 4K isn't even 1%.

We enthusiasts like to think "max settings or GTFO", but the reality is most people are more than willing to turn down settings to get the game running smooth. My cousin still plays all the latest games on a GTX760, including PUBG. He just turns down the settings. The good thing is, most modern games still look pretty decent at low settings.

Most casual gamers, and lets face it Steam is becoming a casual gaming platform, don't mind turning down settings. And with apps like GeForce experience automatically adjusting the settings for them and a lot of games auto adjusting when first run(or when they detect a hardware change), a lot of people don't even realize it. They just fire up the game and play it. Steam attracts a lot of casual console gamers to the PC side.

There are plenty of users with 1080tis and yet still doesn't show, weird...

Obviously not enough to get it un-lumped from the "Other" category. And I'm not surprised. The 1080 is only accounting for 1.4% of users, so I doubt the 1080Ti is even at 10% of that. Despite what most believe, high-end hardware doesn't sell in huge volumes. I thought I read somewhere that each lower price point tends to outsell the next higher price point about 10 to 1. So the 1080 outsells the 1080Ti about 10 to 1, and the 1070 outsells the 1080 by about 10 to 1, and the 1060 outsells the 1070 by 10 to 10, etc.

 

[kruk]

The Steam platform might be the the most widespread digital game sales platform, but it cannot be used as a valid metric for global gamers hardware usage like some people here think. Why? Because the population does not represent the real world scenario, any sample of people you will pick, will be representative only and only of Steam gamers. I do think they are doing a great job sampling, but outside Steam these stats are worth nothing ...

 

[PLAfiller]

The Steam platform might be the the most widespread digital game sales platform, but it cannot be used as a valid metric for global gamers hardware usage like some people here think. Why? Because the population does not represent the real world scenario, any sample of people you will pick, will be representative only and only of Steam gamers. I do think they are doing a great job sampling, but outside Steam these stats are worth nothing ...

I kind of disagree with this one. For a free statistics it is quite something. Of course there are a few "iffs" you need to take into consideration, but their sheer access to a large amount of people is impressive. Organizing a statistic poll from an agency on national level or even global level will require organization, lots cash and results won't be freely available. And here we have quick and dirty data we can work with, presented in nice colorful graphics I'd totally use it to speculate if I want to say something on the matter.

 

[Supercrit]

PC gaming full of peasants confirmed.

 

[Manu_PT]

PUBG is a demanding game, you need to have a really good PC to run it properly, I agree that the other 2 are not demanding but that does not mean that automatically most of the players that are playing those games have low to mid end hardware

Nop you dont. You run pubg on a i3 6100 + gtx750ti at medium settings 1366x768 easily. (90fps+)

 

[Fluffmeister]

PC gaming full of peasants confirmed.

It's comforting... they may be poor, but they aren't stupid.

 

[cdawall]

There are plenty of users with 1080tis and yet still doesn't show, weird...

There are many less users with those vs the 1/7th cost 1050ti

 

[Keullo-e]

Steam breaks down cores as "CPU's". So presumably I guess you are referring to i7 and asking where are they? 0.49% according to the survey.

Edit: ahhhh, never mind. That wouldn't explain hyperthreaded dual cores. They must be mixed in the 4 CPU.

I'm not sure. Steam system information says 2 physical and 4 logical cores.

 

[notb]

There are plenty of users with 1080tis and yet still doesn't show, weird...

Plenty like what?
There are plenty of Mercedes S-class owners, but you don't bump into one on every parking.
It's actually a surprise that 1080 is on this list. 1080ti is way to expensive to get ~1.5% share.

The Steam platform might be the the most widespread digital game sales platform, but it cannot be used as a valid metric for global gamers hardware usage like some people here think. Why? Because the population does not represent the real world scenario, any sample of people you will pick, will be representative only and only of Steam gamers. I do think they are doing a great job sampling, but outside Steam these stats are worth nothing ...

As for "worth nothing" - I disagree.
But you are generally right that Steam is not very representative for the whole population. Of course it will be dominated by people with better gear (and possibly gaming a lot more than average).
Just the fact that 1060 is on the top spot and IGP is 20th shows this is the case.

 

[Manu_PT]

The Steam platform might be the the most widespread digital game sales platform, but it cannot be used as a valid metric for global gamers hardware usage like some people here think. Why? Because the population does not represent the real world scenario, any sample of people you will pick, will be representative only and only of Steam gamers. I do think they are doing a great job sampling, but outside Steam these stats are worth nothing ...

So the other gamers are on.... Blizzard´s battlenet for WoW, Riot launcher for LoL, Origin? with only 20k peak per day playing its most played game (BF1 -> bf1stats.com).

Nah, I disagree with you. To be fair if we could have stats from Riot Launcher users, the percentage of players using low end hardware would be even bigger. LoL is the most played PC game and it runs on a 5 year old Laptop with a HD5000 IGP, and with ease.


Also guys, don´t forget GTX1060 is now the same GPU on both desktop and mobile. So, previously you would have half of its percentage on the Desktop GTX 1060 and the remaining on a "GTX 1060m", 2 different gpus on the stats. Since pascal put the same chips on the laptops, it all counts as one gpu.

If you look at top 20 most used GPUs you find Intel HD5500, GT630, 940m, gtx650, gtx950m, gtx750, gt720m, gt730, gtx950. These chips are really really low-end, and that explains it all.

 

[kruk]

So the other gamers are on.... Blizzard´s battlenet for WoW, Riot launcher for LoL, Origin? with only 20k peak per day playing its most played game (BF1 -> bf1stats.com).

I'm pretty sure the number of PC gamers is a way over a billion, thus Steam represents only a fraction of them (probably 10 % or less). Some people don't use digital sale platforms, some play casual or flash games - but they are still gamers ...

 

[INSTG8R]

Still waiting for Valve to come clean on their "Software & Hardware Survey"
They've never explained exact methods they use to get their results - is it random pickings who gets to survey each month?
It can't be everyone, since the survey doesn't pop up on hardware changes.
What about users who use more than one machine on same account?

Also, they still haven't cleared up obvious mistake that has been in their video cards survey for years.
No - "HD 8800 -series", which is one model sold for one year on prebuilt OEM-machines is not and has never been used more than HD 7800 -series, which includes several models sold for a year in prebuilt OEM-machines and several years on retail.

Actually it popped up on me yesterday. You can do it anytime you like it’s under the Help tab.

 

[Melvis]

These results have to be taking with a grain of salt but its interesting to see none the less.

AMD CPU's have been selling like hot cakes but are not getting used for gaming obviously or the results would be different I would imagine since sales have gone up like 25%

AMD GPU's well there sold out everywhere! and again not for gaming as we all know and people are having a hard time getting a AMD card for gaming so again the results dont reflect that.

OS well again it shows what? how many people actually use or have steam installed which is like less then 30% of all users? and W7 is still the most used OS over W10, again these results dont reflect that.

In the gaming world yep cool thats great, in the none gaming world its the complete opposite.

 

[notb]

AMD CPU's have been selling like hot cakes but are not getting used for gaming obviously or the results would be different I would imagine since sales have gone up like 25%

Sales measured in revenue have gone up 25%, but sales in units might have dropped, actually. So the fact that AMD share is decreasing among Steam users is not surprising.
BTW: it's almost certain that AMD CPU market share among gamers is much higher than among PC owners as a whole.

AMD GPU's well there sold out everywhere! and again not for gaming as we all know and people are having a hard time getting a AMD card for gaming so again the results dont reflect that.

At this point there really is no point in buying an AMD GPU for gaming (maybe other than being very used to FreeSync), so again: drop is not surprising.
Just look how many AMD GPUs on this list are from 2012-2013 (HD7000-8000).
http://store.steampowered.com/hwsurvey/directx/
The oldest GPU in NVIDIA's top10 is a 750Ti from 2014.
The oldest GPU in their top20 is GTX660 - still younger than HD7800.

OS well again it shows what? how many people actually use or have steam installed which is like less then 30% of all users? and W7 is still the most used OS over W10, again these results dont reflect that.

This is irrelevant. Putting mining aside, GPUs are made mostly for gaming, so Steam users can be considered fairly representative in this statistic. This is not true for OSes, obviously.
According to StatCounter, current W7's usage share is around 39%, W10 is fairly close at 31%. But you have to take into account that W7 still dominates in corporate environments.
It's not surprising that W10 is already dominant among Steam users.

 

[notb]

I would love to see such statistics adjusted by actual gaming time (since Steam gathers these data as well).
Beside the desktop, I've also installed Steam on my cheap 2-in-1 (Intel Atom) and I'm pretty sure I've also submitted it's data to Steam. I doubt I've every played anything Steam-operated on it.

 

[Chaos]

At this point there really is no point in buying an AMD GPU for gaming (maybe other than being very used to FreeSync), so again: drop is not surprising.

Yeah, let's ignore the fact that RX x70 and x80 cards were highly competitive with 1060 3/6GB.

 

[ShurikN]

AMD GPU's well there sold out everywhere! and again not for gaming as we all know and people are having a hard time getting a AMD card for gaming so again the results dont reflect that.

I'm guessing the majority of miner cards (polaris ones) are in machines without steam installed.

 

[Toothless]

Quad core, else Thanks, aleast 6 Core system today... the System is alot more smooth overall.

Yeah no, don't even start with the "6+ core" high horse bs since 99% of us here already know better.

 

[kruk]

Yeah no, don't even start with the "6+ core" high horse bs since 99% of us here already know better.

I really hope you run that i7 of yours with hyperthreading disabled ...

 

[Toothless]

I really hope you run that i7 of yours with hyperthreading disabled ...

 

[kruk]

Good. Now disable those 4 additional threads or stop bashing people who say 4 cores is not enough. Thanks!

 

[Toothless]

Good. Now disable those 4 additional threads or stop bashing people who say 4 cores is not enough. Thanks!

I'm not even bashing when you're the one getting all upset over a four-cored processor being enough. i5s are perfectly fine for gaming and will be for a good while more. Even i3s are just fine.