December Steam Survey Numbers Points Towards Slow Death of the Quad Core Gaming CPU

[TheLostSwede]

This might not come as a real surprise, but in the latest Steam hardware survey, we're seeing clear declines of quad core CPUs, the category that some people have been claiming for the longest of times, is all you need for a gaming PC. Among the Windows systems, the decline is over a percent, with six core CPU gaining well over a percent, although the numbers vary quite a bit over the past five months, which is all the history Valve provides. The decline is also clear on OSX, although it's not quite as big percentage wise, but here the biggest growth is in the eight and 10 core segments, most likely due to Apple's introduction of its own M1 variants of CPUs. Only in the Linux segment are the dual and quad core CPU segments increasing, which suggest that some of these systems might be repurposed Windows machines.

The six core and higher CPU segment now holds over 50 percent share in the Steam survey and eight core CPUs are also up somewhat for Windows machines. Of these CPUs, Intel is holding a 69.27 percent share, up 0.82 percent compared to November, although still down over 3.5 percent since August versus AMD. Intel also gained 0.33 percent of Linux users and is back over 60 percent for the first time since August. On the OSX side of things, Apple seems to have gained a 27.97 percent share of Steam users surveyed, up from 6.05 percent just a month earlier. There has also been a 1.2 percent increase in Steam users surveyed that have 16 GB of RAM, suggesting that the low RAM prices in 2021 has made people upgrade their systems. Over 47 percent of all Steam users that were surveyed appear to have at least 16 GB of RAM in their systems.




To no one's surprise, the most popular GPU is the GeForce GTX 1060, which holds a 7.92 percent share, up 0.34 percent since November. Nvidia is without a doubt the most popular choice among Steam users with a 76.83 percent share, with AMD in a slight decline in December with 14.45 percent, with Intel also in a slight decline at 8.51 percent. Apple has a mere 0.77 percent of the total GPU's surveyed, placing them in the other category. This simply suggests that not a lot of Steam users own one of the new Macs or have participated in the survey so far. Among VR headsets the Oculus Quest 2 keeps gaining market share and now holds a 39.62 percent market share among the surveyed Steam users, up 3.3 percent since November. That said, VR headsets are still not all that particularly popular, as only 1.93 percent of Steam users that responded to the survey have one, although it's still a slight increase since the previous month. A quick look at monitors suggest that 1440p is slowly gaining popularity, with an increase of 0.79 percent, but 1080p still holds a solid 67.12 percent in the Steam survey.

View at TechPowerUp Main Site

 

[bug]

The title is misleading. It's not the gaming quad core that's dying, it's the quad core in general. And not because it's not good enough for gaming, it's dying because it's not being made as much as it once was. I mean, we've had 6 cores for ~$200 for how many years now?

 

[PerfectWave]

Seriously still ppl believes at steam survey? LOL

I got steam survey on my vm machine lol

 

[Mussels]

The Steam HW Survey needs to be outdated:

• Why there's only Intel CPU speeds? What about ADL CPUs which have cores with different frequencies? What about AMD CPUs? Haven't they got a frequency? :-D
• System RAM: are you sure there are still people with less than 4GB of RAM? What about thousands with 24/32/48/64GB of RAM? All more than 16?
• Physical CPUs? Why not CPU sockets for good measure? Maybe show "Cores" instead?

You know you can... click and get those details? Physical CPU's is based on how many full cores are reported to the OS, because of the scandals in the past from the CPU makers disagreeing on what defines a core

 

[TheLostSwede]

The title is misleading. It's not the gaming quad core that's dying, it's the quad core in general. And not because it's not good enough for gaming, it's dying because it's not being made as much as it once was. I mean, we've had 6 cores for ~$200 for how many years now?

You know as well as I do that "you don't need more that four cores for gaming" is still going around and many people are still not recommending anything more if you're only going to use your PC for gaming.
It's obviously not true for many modern games, but hence the title, if somewhat misleading when considering the CPU options out there.

Seriously still ppl believes at steam survey? LOL

I got steam survey on my vm machine lol

It's one of few available data points with a significant number of entries. It's just an indicator of what's going on in the market and should by no means be taken as gospel.

 

[Mussels]

Oh and as for the cores:
I've been playing starcraft II, a directX 9 game - and then alt tabbing to use TPU and watch crunchyroll.

While the average is skewed lower due to spending far more time idling, i saw 6.8 of 8 cores used while in a poorly threaded title... and never saw 7 or 8 cores get touched, period - over 5 hours of gaming, web browsing and media watching simultaneously.
(Current, min, max, average)

 

[bug]

You know as well as I do that "you don't need more that four cores for gaming" is still going around and many people are still not recommending anything more if you're only going to use your PC for gaming.
It's obviously not true for many modern games, but hence the title, if somewhat misleading when considering the CPU options out there.

Yes, more games have learned how to use more cores. Still, it's not why the 4-core CPU is going the way of the dodo.

I would go as far as saying the Steam survey as a whole has little to do with games. Whether you're looking at CPU, GPU memory size or resolution, mainstream parts are always in the lead, making the survey more of list of stuff you can buy for ~$200-300.

 

[Fluffmeister]

I see the RX 6X00 cards still barely make a dent too

 

[LemmingOverlord]

I wish everyone had Steam installed on their servers with the survey option on.

That 20 core-40 thread, 128GB Xeon setup would really throw things off on a tangent.

 

[kruk]

It's just an indicator of what's going on in the market and should by no means be taken as gospel.

The problem with Steam survey is that a) Prebuilts and Laptops are overrepresented b) not everyone uses Steam c) Methodology is not transparent.

An example weird thing for this months results: DirectX 8 GPUs and below gained almost 1% of share and that is a category consisting of 20+ years old GPUs. How does that make sense? Also, there is a GPU named AMD Radeon Graphics that appeared in september. What is that?

 

[TheLostSwede]

The problem with Steam survey is that a) Prebuilts and Laptops are overrepresented b) not everyone uses Steam c) Methodology is not transparent.

An example weird thing for this months results: DirectX 8 GPUs and below gained almost 1% of share and that is a category consisting of 20+ years old GPUs. How does that make sense? Also, there is a GPU named AMD Radeon Graphics that appeared in september. What is that?

I think you mean 9, not 8, as there's no 8 in any list I can find at least.
It's obviously quite funny seeing SiS at the lead of any kind of list today, since they haven't made anything graphics related for over a decade that I know of at least.
And yes, there's no transparency and we don't even get to know how many people are being surveyed every month, so it has to be taken for what it is.
However, there are no other data points at all and it is at least something of an indicator of what people are using in terms of hardware.
Steam does have some 23 million plus users online right now, peaking around 27-28 million daily, so it's not an insignificant data point, even if one should be careful with drawing too many conclusions from the data presented.

https://steamdb.info/graph/

 

[TheinsanegamerN]

You know as well as I do that "you don't need more that four cores for gaming" is still going around and many people are still not recommending anything more if you're only going to use your PC for gaming.
It's obviously not true for many modern games, but hence the title, if somewhat misleading when considering the CPU options out there.

Why do you consider it misleading? As Techspot clearly showed in their i3 test last year, a quad core was sufficient to hit over 60 FPS for 1% lows in every game tested, and short of using RTX 3080 class hardware you're stil going to be GPU limited.

Intel 5th-Gen Core vs. 10th-Gen Core Architecture Benchmarked

You'd be forgiven if you had no idea Intel had a 5th-gen desktop CPU range, there were only 2 parts, and neither were on sale for long....

www.techspot.com

Then there is this small issue of the new quad core i3 beating a ryzen 3600 in games:

Intel Core i3-12100 Review: The New Ultra-Budget Gaming King – Art of PC

artofpc.com

The quad core isnt dead yet. In the nursing home yes, but it's still got a few years left in it.

It's one of few available data points with a significant number of entries. It's just an indicator of what's going on in the market and should by no means be taken as gospel.

It's not perfect, but the "wah steam is so terrible it should be ignored" comments are getting annoying.

 

[kruk]

I think you mean 9, not 8, as there's no 8 in any list I can find at least.

If you click on https://store.steampowered.com/hwsurvey/videocard/ there is DirectX 8 GPUs and below in the top section.

However, there are no other data points at all and it is at least something of an indicator of what people are using in terms of hardware.
Steam does have some 23 million plus users online right now, peaking around 27-28 million daily, so it's not an insignificant data point, even if one should be careful with drawing too many conclusions from the data presented.

It definitely is a great source of HW info for Steam users, but that is it. It's impossible to generalize it to PC gaming as a whole, because it's a bubble. Here is a fun read on why such surveys are problematic: https://www2.math.upenn.edu/~deturck/m170/wk4/lecture/case1.html

Much more trustworthy source is: https://www.jonpeddie.com/

 

[bug]

The problem with Steam survey is that a) Prebuilts and Laptops are overrepresented b) not everyone uses Steam c) Methodology is not transparent.

An example weird thing for this months results: DirectX 8 GPUs and below gained almost 1% of share and that is a category consisting of 20+ years old GPUs. How does that make sense? Also, there is a GPU named AMD Radeon Graphics that appeared in september. What is that?

It makes perfect sense. At very low numbers, the random selection for surveys will do that. And it's very likely owners of such old systems don't game each and every month, another variation that will play tricks with outliers.

Fwiw, there's a reason statistics recommend ignoring everything that's outside 3-sigma

 

[TheLostSwede]

If you click on https://store.steampowered.com/hwsurvey/videocard/ there is DirectX 8 GPUs and below in the top section.

Ah, missed that, I was scrolling down the list looking for a header. That one is truly odd, as it's comparatively a massive increase. Could it be the new Apple GPUs that are detected badly?

Much more trustworthy source is: https://www.jonpeddie.com/

Not really. It's not as if they are disclosing much, unless you pay up of course.

 

[newtekie1]

This isn't really a surprise when you consider the i5 has been 6-cores for like 5 years now.

And the fact is you do only need quad-core for gaming. That doesn't mean you don't benefit from something more.

 

[Dr. Dro]

I would argue that Steam's survey is relevant and useful, if not one of the most accurate, even if the sample size is not universal (i.e. every single Steam user), because end of the day, the thing in common between these are systems that they all have Steam installed and thus are computers that game developers have a subjective interest in supporting (otherwise they cannot sell games).

The inaccuracies related to older hardware are more or less to be expected, it is probably bogus/obsolete data. Valve should trim these sometime, as Windows no longer supports any of these older GPUs, you would be hard pressed to run even the lightest desktop environments for Linux on something like SiS Mirage graphics (we are talking early 2000's integrated chips here), and Steam no longer runs on Windows XP or Windows Vista anyway. They should also discontinue Windows 7 support soon, as the OS is no longer maintained and increasingly a security risk. I suspect that will come once Chromium no longer works on 7, which was postponed due to the pandemic but should be soon.

Either way, once the Steam Deck ships, the share of quad-core processors should balloon up again, as should the relatively meager share of Linux users.

 

[dirtyferret]

You know as well as I do that "you don't need more that four cores for gaming" is still going around and many people are still not recommending anything more if you're only going to use your PC for gaming.
It's obviously not true for many modern games, but hence the title, if somewhat misleading when considering the CPU options out there.


It's one of few available data points with a significant number of entries. It's just an indicator of what's going on in the market and should by no means be taken as gospel.

There is one thing you need to understand swede, there are three types of people in this world; those who do understand the basic concepts of math and statistics and those who do not.

You know as well as I do that "you don't need more that four cores for gaming" is still going around and many people are still not recommending anything more if you're only going to use your PC for gaming.

I'm sure it is but I personally don't see it as much. Then again I have not spent as much time replying to threads the last few years as I have done in the past. I've been taking a page out of @EarthDog book and lurking in threads, shaking my head, getting a good laugh, and moving on.

making the survey more of list of stuff you can buy for ~$200-300.

So the items that are the most affordable and give the best bang for your buck performance are well represented?

The problem with Steam survey is that a) Prebuilts and Laptops are overrepresented b) not everyone uses Steam c) Methodology is not transparent.

A) So the two most sold types of PCs are over represented? (FYI, it's two words not one)
B) True but the vast majority of gamers do
C) The survey is actually more for game developers but I appreciate the fact STEAM does release those figures to us. The technique is called random sampling but they do not release the methodology to the public which is fine.


I personally love the STEAM hardware survey but I also took stats in HS and college so I understand it as opposed to many people here who clearly lack basic math skills. I also love the fact that so many people have complete hissy fits because the PC gamer industry (one that includes, tablets, laptops, pre-builts) does not match their niche PC builds of an enthusiast forum.

In fact, I think I'm going to the consumer reports forum and tell them their top ten best selling car/truck list is complete bullshit! I didn't get a survey card this year from CS! Ford, Honda, and Toyota are clearly over represented! There are way too many cars that cost between $20-30k on their list! I belong to a luxury sports sedan forum and I did not see a single BMW M3, Audi S4, or even Mercedes CLS class, WTF consumer reports!

 

[bug]

So the items that are the most affordable and give the best bang for your buck performance are well represented?

Yes. Steam survey tells you those who actually build a gaming system are a tiny minority. Most people game on whatever hardware is mainstream at any given moment.
Basically, you can disregard the survey entirely and just guess who's in the lead.

 

[dirtyferret]

Steam survey tells you those who actually build a gaming system are a tiny minority.

100% true but you and everyone else here already knows this, you don't need the STEAM survey for this

Most people game on whatever hardware is mainstream at any given moment.

Most people game on whatever they can get their hands on at any given moment.

Basically, you can disregard the survey entirely and just guess who's in the lead.

100% true on the former, the STEAM hardware survey has nothing to do with your build or my build or anyone else here. We all have various reasons for our builds but I'm not following on the latter? What lead are you talking about? That Intel and Nvidia sell more CPUs & GPUs than AMD? We already knew that. That after years of having hex core + CPUs on the market, the share of quad core CPUs has been shrinking each month? We need a survey to tell us that?

The STEAM survey tells us one thing; based on a random sample of the most popular PC gaming platform, these are the results. Now STEAM is so large that it can accurately reflect the industry as a whole in the same way Ford and Toyota are the two top selling car manufactures in the USA and if I look at their sales I can accurately reflect the USA car industry.
That doesn't mean you and I drive a Ford or Toyota (or should even buy one) but if the top selling Ford autos are all trucks and three of the top four selling Toyotas are SUVs and a Truck, I can take a random sample of 3k car owners across the USA and have a strong certainty trucks and SUVs will be the most popular autos in that survey and across the country.

 

[kruk]

@dirtyferret: Why an Ad hominem attack? Why not provide sources for your claims and discuss this in civilized manner? I would love to come to the bottom of this "Is Steam survey good or not" once and for all, but someone always comes along and starts raging and insulting and so on ... thanks for wasting another opportunity for this.

P.S: statistics are not basic math skills - proper sampling and results interpretation are extremely hard and you should know this ...

 

[Franzen4Real]

there are three types of people in this world; those who do understand the basic concepts of math and statistics and those who do not.

I've been taking a page out of @EarthDog book and lurking in threads, shaking my head, getting a good laugh, and moving on.

It is unfortunate though, I do miss the thoughtful commentary and perspectives from some of the members that have taken this path. I certainly valued Earthdog's insight in the comment sections.

I think Steam survey is still an interesting data point, and more than likely useful to game developers. I don't know that it is really something meant to be used for market wide sales figures. So context is important when trying to use the numbers for a particular argument.

That said, VR headsets are still not all that particularly popular, as only 1.93 percent of Steam users that responded to the survey have one

but...but... estimates have shown it to have outsold Xbox in 2021... (I know I know, context... global pandemics, silicone shortages...etc etc)
But really, Quest 2 has sold in the 8mil unit range this last year, not bad! The issue with Steam survey is that it can not include standalones which are a much larger user base than tethered PC VR.

 

[dirtyferret]

@dirtyferret: Why an Ad hominem attack? Why not provide sources for your claims and discuss this in civilized manner? I would love to come to the bottom of this "Is Steam survey good or not" once and for all, but someone always comes along and starts raging and insulting and so on ... thanks for wasting another opportunity for this.

P.S: statistics are not basic math skills - proper sampling and results interpretation are extremely hard and you should know this ...

1. I did not to intend a direct attack towards you and apologize if you felt this way.
2. It is extremely frustrating discussing the STEAM hardware survey every single time a thread is posted about it, almost on a monthly basis. I also worked in this field for a long time; Nielsen ratings, MRI market index, Arbitron (now Nielsen audio), quantitative and qualitative analysis, etc., etc., so the countless posts of people not understanding what the numbers say or think it's voodoo magic can get rather tiresome.
3. If you want a civilized discussion feel free to send me private message but I stopped trying to have an education discussion about the STEAM survey years ago. Too many people simply toss a hissy fit because their favorite brand is not at the top or that "inferior" hardware is too well represented. What either one has to do with anyone's PC and how they take that as an insult to their ego is beyond me and maybe that should be the real discussion.

I think Steam survey is still an interesting data point, and more than likely useful to game developers. I don't know that it is really something meant to be used for market wide sales figures. So context is important when trying to use the numbers for a particular argument.

From your virtual lips (fingers) to other's virtual ears (eyes).

 

[GreiverBlade]

well my first 6 core ... my FX6300 was according to ... well you know various sources, a 3 core, although it was not so bad, oh well ... spent enough time on quads and once i switched to my current R5 3600 it was night and day ...

i can actually truly multitask, like right now playing Skyrim SE (non AE) while posting that (lot of tabs open tho) and running MSI APP player for autofarming in Azur Lane while listening to some music on Spotify and browsing images ... on my 6600K if i did that the emulator would be in "slideshow" mode and various slowdown here and there


nonetheless "Steam survey" i considered that one as much as Windows WEI back in the days .... but now "it's fine" i guess.

 

[Vayra86]

You know as well as I do that "you don't need more that four cores for gaming"

Really? Do those guys still run Windows XP, too? I mean that really is 2015-2017 (PS4 era) at the latest-news, right there. Quads without HT have been performing admirably and increasingly shit since then. I saw my i5 drop more and more frames as the years went by.

The STEAM survey tells us one thing; based on a random sample of the most popular PC gaming platform, these are the results. Now STEAM is so large that it can accurately reflect the industry as a whole

This. And the vast majority should re-read this a few dozen times to get it memorized.

We should take this as literally as we can. A huge number of PCs have Steam installed. It does reflect the industry as a whole - the one of 'PCs that can install Steam'. The problem you describe when it comes to looking at the Steam Survey is that there is absolutely no context whatsoever to these numbers. Its 'an RNG slice' that should represent the big numbers. And that's it. You don't know what each segment of that PC market is spending, you don't know what it plays, you don't even know if they visited Steam for anything other than clicking through the survey box.

Data without context is just that. Data. Trying to add context ourselves that is anything other than 'hey, that looks credible' is a fool's errand. All you can gather from it is the major shifts in the market, long-term, as in 3-5 year scope. Basically what common sense already dictated without looking at numbers.

I think Steam survey is still an interesting data point, and more than likely useful to game developers. I don't know that it is really something meant to be used for market wide sales figures. So context is important when trying to use the numbers for a particular argument.

The Steam Survey we get to see certainly isn't an interesting data point, but what Valve can offer to developers might very well be enriched in various ways. The big one being spending behavior. The pessimist might even think we're providing free lunch for Valve here. And they are correct We get a carrot and post it here, but the carrot is hollow.