Steam Paints A Different Picture Of Intel, AMD, And Nvidia

[LAwLz]

4 minutes ago, NvidiaIntelAMDLoveTriangle said:

Wasn't steam taking the information from all its users without any questions?

I remember reading somewhere some years ago, that steam their surveys are just collected data from everyone who uses steam, and no questions asked of its users.

You're probably thinking of when they were looking through peoples' DNS caches to see if they had visited websites that sell CS cheats.

They only did that to a few people too.

[Arika]

Jesus Christ this thread is cancer.

 

Why can't people just be happy with what they have and stop trying to say "x is better than y because z" if it works GREAT, have fun with it. but stop trying to shit on the other company because it's different that what you bought. you don't need to justify your purchase to anyone.

 

The fanboy wars that have been reignited by the ryzen release (and soon to be Vega) are the worst thing to come out of the hardware market in recent years. both sides are trying to make justifications as to why theirs is better. Hell Vega isn't even out yet and the AMD fans are jerking their e-peen about it without even knowing how well it's going to perform and the Nvidia fans are doing the same thing with volta.

 

The only people who should care about market share are the manufacturers and stakeholders, not the consumers.

[Gravesnear]

24 minutes ago, LAwLz said:

 

Also, if you think Nvidia spent the time and effort to enable DirectX12 on Fermi just so that their Steam survey statistics, which only shows up if you go into the "more info" page and manually summarize the list of graphics cards, would go up then you're quite delusional.

Especially since Fermi has such a low market share that it doesn't even show up in the detailed list of video card usage. 0 Fermi cards appear because their market share is too low, and only two Kepler cards appear (550 Ti and 560), and they total a whopping 1.1% of all cards in the survey. But yeah, I'm totally with you that Nvidia did it just so that they could increase their market share.

 

Thanks for the good laugh though. It's always fun reading what inane comments people post in AMD vs Nvidia/Intel threads. It's like looking through a portal to a whole different dimension where logic ceased to exist.

 

If you actually read the source past page 1, seeking alpha specifically states they sorted GPUs by DX12 support to understand adoption of current generation hardware. I may in fact be wrong - I don't know the full capabilities of steam survey - but that is info THEY use on their finance journalism website. Though I did state that the impact would not be large. I doubt many people are still using those space heaters, but even half a percentage point is important when talking usage share. It is not delusional to think that any company would do this. It is foolish not to. As for the first paragraph in your quoted text above, why else would they do it? DirectX 12 is so poorly implemented on Fermi that it almost doesn't run. That's coming from Tech Report.

 

Logic DOES exist, when you actually read sources and follow tech news by actual journalists.

 

So don't assume I'm an AMD fanboy, you can see the graphics card I use in my profile (gtx 1060). I happen to follow the industry closely because I trade in these companies. Stop writing people off.

[LAwLz]

11 minutes ago, Gravesnear said:

If you actually read the source past page 1, seeking alpha specifically states they sorted GPUs by DX12 support to understand adoption of current generation hardware.

Yes, but it is sorted by graphics card type and not vendor. Also, it is not summarized like the statistics on the first page. So anyone wondering what percentage of DX12 capable cards are from Nvidia vs from AMD will have to go through and add things together by hand. Also, like I said, there are so few Fermi cards still in use that it doesn't even show up in the statistics of the Steam survey. So no, the Fermi update did not change anything and even if it did, it would only show up if someone bothered to go into the detailed view of the statistics and manually calculate the market share by hand.

 

14 minutes ago, Gravesnear said:

I don't know the full capabilities of steam survey - but that is info THEY use on their finance journalism website.

Well it's a pretty bad website in that case.

 

15 minutes ago, Gravesnear said:

I doubt many people are still using those space heaters, but even half a percentage point is important when talking usage share. It is not delusional to think that any company would do this. It is foolish not to. As for the first paragraph in your quoted text above, why else would they do it? DirectX 12 is so poorly implemented on Fermi that it almost doesn't run. That's coming from Tech Report.

1) The market share doesn't change at all actually. What will change is their market share if you are looking at a subsection of the market, and that's at the expense of another subsection. Those subsections are not even important enough for Steam to automatically calculate.

2) Why would they implement it? Because they promised to do so in a PR move a long time ago. In fact, the tech report article you are likely referencing gave that as the reason why they probably added it without a big announcement.

3) The Tech Report article you are most like referencing does not really say DX12 is poorly implemented. It says it lacks support for a lot of DX12 APIs (which is understandably since it was released along time before DX12 was finished) but it also says:

Quote

The GTX 560M is a mobile part based on GF116—a six-year-old GPU at this point—and frankly, the fact that it can complete the benchmark at all is impressive. Hats off to Nvidia for continuing to support truly ancient hardware.

 

What you claim they said: "DirectX is so poorly implemented in Fermi that it almost doesn't run"

What the article says: "The fact that it can complete the benchmark at all is impressive"

 

I am not saying that you're an AMD fanboy (especially not after you have said that you're economically invested in this), but I will say that you seem to not tell the entire truth and are making (in my opinion) silly conspiracy theories.

[Jito463]

2 hours ago, zMeul said:

that's not how it works

AMD has already a small market share CPU wise

and for AMD to surpass Intel in overall market share they would've needed not only to sell shit ton more than they usually do, they would've needed to sell more than Intel sold for years - and that is just not possible since Ryzen launched, 3-4 months was it

AMD doesn't have to surpass Intel to increase market share, they just have to increase their current sales faster than Intel increases theirs.  It's growth of sales - not total sales - that determines market share fluctuations.

[Gravesnear]

27 minutes ago, LAwLz said:

 

I am not saying that you're an AMD fanboy (especially not after you have said that you're economically invested in this), but I will say that you seem to not tell the entire truth and are making (in my opinion) silly conspiracy theories.

I'm just saying that's how seeking alpha is deriving its information. What I said was in good faith, without malice. It's a discussion and your opinion is valid. Like I said I could be wrong. I do think it's unfair to call me delusional after I used the information given to me to form an opinion. You could have simply disagreed and stated why, instead of making assumptions on where my head's at.

 

Also it's not really a conspiracy theory or even fringe theory. Millions of dollars are made and lost every day because of small tweaks to data like this shown to people without context. It happens so often, I'm not even surprised any more (i just wish I could get a piece).

 

Still. If what you're saying is true, you are probably right, and I am probably wrong. I acquiesce on that front but still insist that you were kind of a jerk.

 

Edit: And yes, they're not a great website.

Edit: And the tech report thing I was paraphrasing. They literally said it was impressive that it ran at all. I did not use quotation marks when I said it almost doesn't run. Don't make it like I grossly misrepresented what they said.

Edited July 11, 2017 by Gravesnear
Addition

[Yoinkerman]

I though Intel integrated gfx has the majority of market share

[LAwLz]

20 minutes ago, Gravesnear said:

I'm just saying that's how seeking alpha is deriving its information. What I said was in good faith, without malice. It's a discussion and your opinion is valid. Like I said I could be wrong. I do think it's unfair to call me delusional after I used the information given to me to form an opinion. You could have simply disagreed and stated why, instead of making assumptions on where my head's at.

 

Also it's not really a conspiracy theory or even fringe theory. Millions of dollars are made and lost every day because of small tweaks to data like this shown to people without context. It happens so often, I'm not even surprised any more (i just wish I could get a piece).

 

Still. If what you're saying is true, you are probably right, and I am probably wrong. I acquiesce on that front but still insist that you were kind of a jerk.

 

Edit: And yes, they're not a great website.

Edit: And the tech report thing I was paraphrasing. They literally said it was impressive that it ran at all. I did not use quotation marks when I said it almost doesn't run. Don't make it like I grossly misrepresented what they said.

Fair enough. I can't really blame you for looking at what was reported on the website. I would argue that the source is paining a rather inaccurate picture though.

 

 

Just now, Yoinkerman said:

I though Intel integrated gfx has the majority of market share

Overall they do, but not when you only look at people who has Steam installed (which is a minority of all PC users).

[Gravesnear]

Just now, LAwLz said:

Fair enough. I can't really blame you for looking at what was reported on the website. I would argue that the source is paining a rather inaccurate picture though.

Fair enough.

[Doobeedoo]

Well I wouldn't take Steam statistics like some world wide graph in general anyway like most do. It's a part from certain platform and survey criteria etc. But it gives a look to some extent. 

As far as AMD I'm sure that with time they will gain more market share with Ryzen and Polaris once prices normalize and Vega released. It just needs time and good sales for statistics to change. 

[Humbug]

Finally starting with the July results Ryzen starts to make an impact

http://store.steampowered.com/hwsurvey/processormfg/

AMD will be hoping that they can keep this up for the next year and see their market share crawl up every month.

[Jito463]

50 minutes ago, Humbug said:

Finally starting with the July results Ryzen starts to make an impact

http://store.steampowered.com/hwsurvey/processormfg/

AMD will be hoping that they can keep this up for the next year and see their market share crawl up every month.

Eh, I take all those statistics with a massive grain of salt.  As Samuel Clemens is often quoted as saying, 'There's lies, d****** lies and statistics'.

[DarkBlade2117]

On 7/11/2017 at 8:04 AM, Scotter97 said:

Just wait, ebay and craigslist will be full of those RX cards when this mining phase comes to a close.  MANY near death cards. (lol)

No proof that mining shortens the cards life. Many and most good miners undervolt their cards and keep them in temperature controlled rooms. Most you'd have to worry about is the bearing on the fans but even then those are meant to last years.

[Phate.exe]

On 7/11/2017 at 7:51 AM, tom_w141 said:

Take the survey with a pinch of salt its opt in and steam has never asked me to. So my results and likely many others are not included.

Haven't been asked for my previous machine or my current one.

[Humbug]

14 minutes ago, Phate.exe said:

Haven't been asked for my previous machine or my current one.

It's a sample.

Not everybody is asked.

[Humbug]

46 minutes ago, Jito463 said:

Eh, I take all those statistics with a massive grain of salt.  As Samuel Clemens is often quoted as saying, 'There's lies, d****** lies and statistics'.

No salt required.

 

Just a basic understanding of what it is

-a limited sample size (opt in too)

-margin of error applies as always in statistics

-not trying to be representative of entire market (only steam customers)

 

It's one more data point. Which we can track over the next few months. That's all..

[Derangel]

49 minutes ago, Jito463 said:

Eh, I take all those statistics with a massive grain of salt.  As Samuel Clemens is often quoted as saying, 'There's lies, d****** lies and statistics'.

Very true. We cant even take the statistics as an example of the over-all Steam user base. We don't know how many users participated in the poll nor do we have any information on the total number of unique users Steam has on a month-to-month basis. We don't even know how many unique users Steam has over-all.

[Prysin]

On 11/07/2017 at 1:51 PM, tom_w141 said:

Take the survey with a pinch of salt its opt in and steam has never asked me to. So my results and likely many others are not included.

you can also opt in multiple times with the same system and just keep stacking results for one vendor.

 

Given how underhanded Intel has been in the past, i would not be shocked if they had some room with scripted systems just stacking shitloads of surveys and logging onto steam to run some canned game that is easy to script to feign activity.

 

There is also the fact that NOBODY knows how long it takes for steam to "remove inactive survey results". There is no evidence of them doing this monthly, quarterly, annually etc... we have NO PROOF. Thus for all we know, AMD could have risen 20% in market share in actual systems, but if the same 20% intel systems "not being used atm" is still in the survey pool, then it makes no fucking difference.

[Prysin]

4 hours ago, Humbug said:

No salt required.

 

Just a basic understanding of what it is

-a limited sample size (opt in too)

-margin of error applies as always in statistics

-not trying to be representative of entire market (only steam customers)

 

It's one more data point. Which we can track over the next few months. That's all..

actually its worse

[Prysin]

4 hours ago, Humbug said:

It's a sample.

Not everybody is asked.

Actually, it automatically does the survey whether you like it or not, the "first time" you install steam. Wipe steam off the system, go reinstall it and it will run the survey parrallell to asking for your broadband speed and preferred network speed cap etc...

[Prysin]

On 11/07/2017 at 5:14 PM, LAwLz said:

Fair enough. I can't really blame you for looking at what was reported on the website. I would argue that the source is paining a rather inaccurate picture though.

 

 

Overall they do, but not when you only look at people who has Steam installed (which is a minority of all PC users).

paining a rather inaccurate picture

how do you do that? 40 lashes in artistic pattern? cutting oneself and painting with blood? Or is this some secret swedish cult practise you have forgotten to share with the world?

[Dabombinable]

8 minutes ago, Prysin said:

Actually, it automatically does the survey whether you like it or not, the "first time" you install steam. Wipe steam off the system, go reinstall it and it will run the survey parrallell to asking for your broadband speed and preferred network speed cap etc...

Still have a bit of a laugh when Steam registers 2x PIII 1000+2GB RAM+20GB HDD+Geforce 6200 512MB.