[JPR] The balance of power in gaming

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Pariah

Elite Member
Apr 16, 2000
7,357
20
81
It's not working. They haven't had a profitable quarter since 2014. When was the last time AMD released an anticipated product that wasn't a disappointment at launch? The rise in stock price means nothing if they can't start producing a consistent profit. With the release of Zen and Vega in the first half of next year, we will know soon enough how much AMD has left in its present form.
 

USER8000

Golden Member
Jun 23, 2012
1,542
780
136
So when Nvidia is using JPR,its AMD PR helping them?? I know AMD PR has been accused of many things,but that is a new one for me! :p

Nvidia has been using JPR since 2005:

https://www.nvidia.com/object/IO_18301.html

Nvidia from 2005 said:
Latest Market Share Numbers from Mercury Research and Jon Peddie Research Show NVIDIA GPU Market Share on the Rise

2007:

http://www.nvidia.com/object/io_1190676485302.html

The use of powerful graphics is not limited to just games anymore. With their new GeForce mGPUs, NVIDIA is pioneering the development of new technologies that are helping to evolve the capabilities of mainstream PC platforms." — Jon Peddie, president of JPR

Nvidia used them this year. An example:

https://blogs.nvidia.com/blog/2016/02/16/vulkan-graphics-api/

“To be able to play a game like The Talos Principle on the same day an API launches, is an unheard of achievement,” said Jon Peddie, president of Jon Peddie Research. “NVIDIA’s multi-platform compatibility and fully conformant driver support across many operating systems is a testament to the company’s leadership role in Vulkan’s development.

Nvidia PR has no problem using JPR. They have been using them for 11 years.

I think Nvidia knows better about what companies they use in their PR releases,than random people on a forum.

Edit to post.

Nvidia used them on Twitter this year too:

https://twitter.com/nvidia/status/715238319966302209

What so special at #GTC16? Jon Peddie @GfxSpeak shares everything he's looking forward to: http://nvda.ly/105Yf7
https://t.co/VMF3acCpSh
 
Last edited:

Pariah

Elite Member
Apr 16, 2000
7,357
20
81
I'm guessing the JPR works for AMD comment was an attempt at sarcastic humor that went over like a lead baloon.
 

USER8000

Golden Member
Jun 23, 2012
1,542
780
136
I'm guessing the JPR works for AMD comment was an attempt at sarcastic humor that went over like a lead baloon.

Looking at one or two of the comments,the posters actually believe it and think it is some AMD driven conspiracy to do try and do some random thing to the stock price.

If AMD PR was that good they would be doing far better than they are currently.

Both AMD and Nvidia use JPR and Mercury Research for PR as they feel fit. This news article has been reported by sites,which according to posters on this very forum either call them anti-Nvidia or anti-AMD both at the same times.

Its hilarious - soon Mercury Research will be accused of trying to push Nvidia stock price up by some posters.

Not everything is a conspiracy just because its the interwebs.

The markets disagree amd is up 10%. (11/17, dated in case future discussion calls for it)
But I'm extremely wary of amd stock. I knew Nvidia would jump up but not by the amount it did. I can only see amd stock falling in the future once people realize that amd is simply unable to be profitable even if the product is good.

Its probably up since they got some wins with Google:

http://www.nasdaq.com/article/amd-soars-again-after-google-cloud-deal-cm710873

The last jump happened after they won a contract with Alibaba.

Interestingly enough one was a dual Fiji card and the other a dual Tonga card.

Its understandable the stock jumping if two very large companies have decided to show faith in AMD professional solutions,when Nvidia has a large percentage of that market.

The Google win was for deep learning too. This is the first time there has been any hint of a company using AMD technology for this area.
 
Last edited:

ocre

Golden Member
Dec 26, 2008
1,594
7
81
Yep, I read it.

They used Steam data for segmentation and differentiation only, for statistics they used their own (JPR) data.

Steam data was used to make these projections. The statistics in these charts are calculated not raw and steam survey data is used in their calculations.

Steam survey is immediately and totally disregarded when it says something that we dont want to hear. I always believed the survey to contain data that is not entirely useless. But certainly, its not an end all be all. Even the gaming market is bigger than steam, so its representation of that market is not definitive. Not to mention that the hardware survey comes from a pop up that is optional. Some are just naturally more likely to go thru with it than others. Which could amount to a tendency for a certain group, which will shift the results.

Steam never claims their data accuracy, as far as I know. That's another problem.
But, as I say these things, I still think there is value in the steam survey results. Its a huge data set and there is value in it. Just know what you are looking at and understand to take caution in reading the results.

One consistent that should be noted is the monthly trends do seem to be consistent, that they do seem to reflect rising and falling over time. Which should be acknowledged, this is a random survey with different users every month. Its data is not completely meaningless.
 

ocre

Golden Member
Dec 26, 2008
1,594
7
81
Im not sure about their numbers, but your analysis is surely wrong.
Adding up shipments and sales can't result in decrease in units. At least until they are not selling negative hardware, like a negative 0.5GB of VRAM ;)

Perhaps I my point wasnt clear. I kind of hate posting on a phone at work, always a sloppy result..

The article is supposed to represent the current user market share. They add up sales over the past several years and count them as individual gamers. They are assuming that these machines are all still being used for gaming and that each one is a separate person.

Counting years old machines that are all assumed to be active gamers and in the tally. See not only do years old gaming laptops break, many gamers buy new hardware to replace their aging hardware. That its strange to assume that all those machines are still being used and counted in their total gamer market share data. I
This report is suggesting these machines represent individuals. I think it's very very fuzzy.

Other issues,
The consoles just now become x86 machines, yet the data they give goes back well before x86 consoles. Its impossible yet there it is, right there in the chart. Did JPR not realize that the last gen consoles were not x86? Or did someone one on their team who put this together not know? Its entire bases is supposed to be x86 gamer share but they have large numbers and data from non x86- last gen consoles.

Also, even if you disregard the flaw of charting millions of non x86 machines in their x86 gamer market article - the numbers do not even add up. The amount on non x86 consoles shown on their chart as x86 machines is higher than the xbox 360 sales. Did they add in nvidia gpu ps3 sales and count them for amd?

The last problem.
Amd notebooks are being counted as a gamer in a very high percentage. Many of them are not viable gaming machines at all. They just don't have they capability. That's not a knock on them, just saying. Many people who bought less power apus didn't buy it as their primary gaming device.
They throw out tablets, even x86 tablets. I know far more people who have tablets that are used to play games than people who use low power notebooks. I would say that nearly all tablets in homes are used to play games. I mean, that is pretty much what tablets are used for in homes. Entertainment, fun, and games. These tablets are gaming devices.
But yeah, throw out tablets and show non x86 consoles. X86 mobile, thrown out.

Gamers have many outlets. And I do think amd is serving a good chunk of these gamer outlets. But these figures are totally manipulate. And for what reason?
Gamers have many devices. They go through many devices. Many frequently update as well. Most ps4 owners were previous console owners as well. Many still have their old console. They may have a netbook and a desktop pc.

But if we decide to take the individual out and say strictly from a device standpoint. The data is so skewed here, its completely bizarre. There are so many huge problems in this, its unbelievable.
My opinion is now leaning towards intentional, its hard to see how its accidental. I have moved on from how to why.
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
But if we decide to take the individual out and say strictly from a device standpoint. The data is so skewed here, its completely bizarre. There are so many huge problems in this, its unbelievable.
My opinion is now leaning towards intentional, its hard to see how its accidental. I have moved on from how to why.

Because we know nobody uses Intel iGPUs at all for any kind of gaming. *sarcasm*
 

krumme

Diamond Member
Oct 9, 2009
5,952
1,585
136
When people buy a ps4 they dont buy an amd gpu they buy a playstation.
When people buy a 1060 they buy a nvidia gforce.
Obviously that is reflected in margins and in said companies profit.

Its nonsense to dicuss how the numbers is made up just because they dont fit some agenda - like the usual brawl about steam numbers. Why is it not possible to just accept the numbers?

In this case the importance is not if the marketshare is precisely 57% clearly tons of consoles is sold.
The important takeaway is amd is dependant on a segment where they in reality have no brand presence. They are selling tap water instead of water in bottles. Its clearly less profitable but its also more fragile and long term risky.
Right now nv is becomming the same as a gfx. Like there was no word for car. Only VW. Its incredible valuable and imo 10 times more important than having your arch in most devices.
 
  • Like
Reactions: xpea
Aug 11, 2008
10,451
642
126
I see only one useful metric for combining the data like this. It is an indication that because they have a strong presence from the consoles, and thus a >50% total share of devices using the same architecture, GCN may be more directly compatible with newer games because of similar architecture to the consoles. Otherwise, it is really a sort of "captain obvious" type of statement that tells very little about the competitiveness (or lack of it) of AMD in the dgpu market.
 

Krteq

Senior member
May 22, 2015
991
671
136
Hmm, market share data for Q3 2016 from Mercury Research

Total Market Share
NV - no change
AMD +0,2%

Desktop Discrete GPUs
nV -2,7%
AMD +2,7%

d9175582_d885af5462868ad04d50c77c69de3d79.png
 

AtenRa

Lifer
Feb 2, 2009
14,000
3,357
136
like the usual brawl about steam numbers. Why is it not possible to just accept the numbers?

The problem with steam is that people take the Steam Hardware Survey numbers for GPU Shipments. Its not, Steam Hardware Survey numbers are only valid for people using Steam.


And let me show you and everyone else why,

Steam Hardware Survey October 2016 DX-12 cards

http://store.steampowered.com/hwsurvey/videocard/
2itn0wo.jpg



Actual Q3 2016 GPU shipments
https://redirect.viglink.com/?format=go&jsonp=vglnk_147972014579911&key=8a27173f1d0514db88df01b1c3d4a370&libId=ivruier80101045j000DAlr9nbvz7&loc=https://forums.anandtech.com/threads/jpr-the-balance-of-power-in-gaming.2492145/page-2&v=1&out=http://www.mercuryresearch.com/graphics-pr-2016-q4.pdf&ref=https://forums.anandtech.com/threads/jpr-the-balance-of-power-in-gaming.2492145/&title=Page 2 - [JPR] The balance of power in gaming&txt=market share data for Q3 2016 from Mercury Research


d9175582_d885af5462868ad04d50c77c69de3d79.png


In Desktop AMD gained share from 22,8% vs 25,5% Q2 to Q3 , while NVIDIA lost share from 77,2% vs 71,2% Q2 to Q3.
In order for AMD to have gained share, they needed to ship more GPU than NVIDIA, do you actually see this AMD increased GPU shipments in the Steam Hardware Survey above ?? There is not even a single Polaris GPU in that Steam survey for DX-12 cards.

Steam Hardware Survey doesnt mean Global GPU Shipments, you may only take the survey numbers for what GPUs people using on STEAM and nothing more. And in that context, without knowing how many people took the survey its even less reliable to use Steam numbers.

Edit: RX 480 it does appear in the DX-12 cards at 0.39% with a first appearance in October.
 
Last edited:
  • Like
Reactions: Bacon1 and Krteq

Dribble

Platinum Member
Aug 9, 2005
2,076
611
136
Steam hardware survey is fine, it just doesn't match the reality distortion field you try to maintain around AMD's success. Accept the truth of where AMD really is right now and suddenly what you see on steam makes perfect sense.
 

krumme

Diamond Member
Oct 9, 2009
5,952
1,585
136
We have jpr and mercury giving us number for marketshare with and without consoles.
They are usually pretty close and who doesnt trust those data to be fairly accurate? Obviously steam in itselves is showing something a bit different but its interesting in another context. Not for marketshare. We have the pro giving us that information.
Jpr, mercury and financial data is what we need.
We are constantly discussing how the numbers is made not what the implication and effect is.
As if its interesting if nv or amd goes up or down a single quarter just looking at the number. I think its a bit childish like my dad your dad sort of thing. Lets move on and discuss the interpretation and consequences.

What is interesting and perhaps a bit surprising from this q estimates from jpr and mercury?
 

LTC8K6

Lifer
Mar 10, 2004
28,520
1,575
126
RX480 4gb at $199 is Amazon's best selling graphic card at the moment. However, there are only two AMD cards in the top 20, both MSI 480 cards. One a 4gb card and the other an 8gb card. The 1050Ti is already at number 3.

I can't recall the last time an AMD card held the #1 spot.
 

[DHT]Osiris

Lifer
Dec 15, 2015
14,004
12,069
146
RX480 4gb at $199 is Amazon's best selling graphic card at the moment.

I wonder what % of those cards are going to actual gamers, and what % ends up in bitcoin mining arrays? I feel like for 'gamers', aka what those steam #'s represent, most people would spend the extra $100 for the double perf or whatever of a 1070.
 

Headfoot

Diamond Member
Feb 28, 2008
4,444
641
126
Steam hardware survey is fine, it just doesn't match the reality distortion field you try to maintain around AMD's success. Accept the truth of where AMD really is right now and suddenly what you see on steam makes perfect sense.
It is not. It's not totally useless but it's not "fine" for the purpose people want to use it for. It's just basic data and statistics 101. Opt-in surveys are significantly less accurate than a true sampled survey. Larger N = better data, etc. Known errors in discovering cards or only one of multiple GPUs taking priority over another (e.g. Intel iGPU getting counted over discrete in some cases) It's all really basic stuff.

It's useful for what it is, a generalized outlook of Steam user's hardware for game developers to plan against. For that purpose it doesn't need to be highly accurate, general trends suffice. When knuckledraggers on the internet want to use it as ammo to cheerlead how their favorite brand is totally destroying the other brand, then it's very much not "fine" for that.
 
  • Like
Reactions: Bacon1

[DHT]Osiris

Lifer
Dec 15, 2015
14,004
12,069
146
It's useful for what it is, a generalized outlook of Steam user's hardware for game developers to plan against. For that purpose it doesn't need to be highly accurate, general trends suffice. When knuckledraggers on the internet want to use it as ammo to cheerlead how their favorite brand is totally destroying the other brand, then it's very much not "fine" for that.

Considering this thread is about the balance of power in *gaming*, I would presume that whatever game developers are planning against is a pretty good indicator of what the current trend is in gaming. I know if I were a game dev, I'd be focusing a lot more on nvidia than on amd right now.
 

USER8000

Golden Member
Jun 23, 2012
1,542
780
136
Considering this thread is about the balance of power in *gaming*, I would presume that whatever game developers are planning against is a pretty good indicator of what the current trend is in gaming. I know if I were a game dev, I'd be focusing a lot more on nvidia than on amd right now.

Considering that consoles are the lead platforms for many games,many devs are actually concentrating on consoles first and if you were a dev,you would be concentrating more on the PS4 PRO and XBox Scorpio. For many devs there is simply more money to be made from consoles as the games cost more and even things like online play are not free. When a company like Blizzard who ignored consoles for over a decade(even though they released titles on consoles in the 1990s),has started to even release games like Overwatch on console and when even Planetside 2 of all games now is on consoles,it its clear to see where the trends are moving towards.

It makes massive sense as spreading costs over multiple platforms simply makes modern game development more viable than putting all your eggs in one basket.

BF1 on consoles is far more optimised than the PC equivalent for example.

Look at some of the most successful PC games like League of Legends,DOTA2,Minecraft,etc - they don't require a high end card to run on. Even integrated graphics will work and that is because even on PC there is more money to be made from targeting the lowest common denominator.

Most of us having decentish PCs are using them to play console ports with better graphics. Very few high budget games in the next few years will be PC only - I can only think of Star Citizen and that is only PC only since PC gamers are throwing money at the dev before the game is even being released.
 
Last edited:

[DHT]Osiris

Lifer
Dec 15, 2015
14,004
12,069
146
Considering that consoles are the lead platforms for many games,many devs are actually concentrating on consoles first and if you were a dev,you would be concentrating more on the PS4 PRO and XBox Scorpio. For many devs there is simply more money to be made from consoles as the games cost more and even things like online play are not free. When a company like Blizzard who ignored consoles for over a decade(even though they released titles on consoles in the 1990s),has started to even release games like Overwatch on console and when even Planetside 2 of all games now is on consoles,it its clear to see where the trends are moving towards.

It makes massive sense as spreading costs over multiple platforms simply makes modern game development more viable than putting all your eggs in one basket.

BF1 on consoles is far more optimised than the PC equivalent for example.

Look at some of the most successful PC games like League of Legends,DOTA2,Minecraft,etc - they don't require a high end card to run on. Even integrated graphics will work and that is because even on PC there is more money to be made from targeting the lowest common denominator.

Consoles tend to take the lead for gamers who aren't generally technically inclined (or don't have the money to do the PC customization thing). Yes, much of gaming is done on consoles (as it always has been) but the requirements for consoles is well documented, and I'm sure quite a bit easier to create for, vs the myriad possibilities with PCs, hence why they *tend* to be better optimized on their specific technical level vs the PC equivalents. Having said all that, the bracket targetted by these sub-$200 card tier tends towards the same group that would likely just buy whatever the most recent playbox720 is, and call it done.

Therefore those gamers who are skipping the console and buying/building a computer, tend to get something at the middle->high end of the spectrum, more in the $250-$400 range. In this space, AMD is a non-starter (and frankly has been for a while).

What people need to do is drop the pretenses, and define what they're looking for to decide whether or not their 'side' is winning. Since this is a red vs green measuring contest, the rules need to be laid out. Is this purely PC gaming video card sales? Are we talking total hardware running on x product? Number of GPU cycles spent rendering video game assets? Phones? Steam charts? Sales over the last 2, 4, 10 years?
 

USER8000

Golden Member
Jun 23, 2012
1,542
780
136
Consoles tend to take the lead for gamers who aren't generally technically inclined (or don't have the money to do the PC customization thing). Yes, much of gaming is done on consoles (as it always has been) but the requirements for consoles is well documented, and I'm sure quite a bit easier to create for, vs the myriad possibilities with PCs, hence why they *tend* to be better optimized on their specific technical level vs the PC equivalents. Having said all that, the bracket targetted by these sub-$200 card tier tends towards the same group that would likely just buy whatever the most recent playbox720 is, and call it done.

Therefore those gamers who are skipping the console and buying/building a computer, tend to get something at the middle->high end of the spectrum, more in the $250-$400 range. In this space, AMD is a non-starter (and frankly has been for a while).

What people need to do is drop the pretenses, and define what they're looking for to decide whether or not their 'side' is winning. Since this is a red vs green measuring contest, the rules need to be laid out. Is this purely PC gaming video card sales? Are we talking total hardware running on x product? Number of GPU cycles spent rendering video game assets? Phones? Steam charts? Sales over the last 2, 4, 10 years?

Its not the gaming - its the development. Consoles are lead platforms for development now. Its not going to change anytime soon. When you look at the financials of some of the major gaming companies,they tend to be making more money from consoles in many cases. So ultimately its consoles which will determine the rate at which technology is accepted - why do you think DX9 lasted so long?? It was because the consoles had GPUs which had hardware features more consistent with DX9 not later APIs. Why do you think BF1 is noticeably better going to 4 and even 8 threads?? Consoles.

Don't get me wrong I would love to see the good old days of somebody like Crytek releasing another Crysis and not giving two hoots about consoles. But look what happened to Crysis and their PC-centric approach. They failed because PC owners bitched and moaned since their £250 card could not max out Crysis.

Now look back at Crysis nearly a decade later - it looks utterly amazing for such an ancient game.

Its not going to happen anytime soon. Just look at more and more of the PC-centric devs like CDPR?? They moved to consoles too,and look at how much The Witcher 3 got graphically downgraded over initial previews??

It was because of consoles and also probably because CDPR probably realised most people would not have a fast enough graphics card to run the game at those kind of settings anyway.
 

bystander36

Diamond Member
Apr 1, 2013
5,154
132
106
The obvious difference between JPR and Steam surveys, is JPR is telling you what is being sold in a particular small time frame. Steam shows what gamers are currently using, which may span many years worth of ownership and doesn't track GPU's being used for other purposes.

I really doubt the opt in and opt out aspect of Steam plays a significant role in results. It just shrinks the data a little.
 

[DHT]Osiris

Lifer
Dec 15, 2015
14,004
12,069
146
Its not the gaming - its the development. Consoles are lead platforms for development now. Its not going to change anytime soon. When you look at the financials of some of the major gaming companies,they tend to be making more money from consoles in many cases. So ultimately its consoles which will determine the rate at which technology is accepted - why do you think DX9 lasted so long?? It was because the consoles had GPUs which had hardware features more consistent with DX9 not later APIs. Why do you think BF1 is noticeably better going to 4 and even 8 threads?? Consoles.

Don't get me wrong I would love to see the good old days of somebody like Crytek releasing another Crysis and not giving two hoots about consoles. But look what happened to Crysis and their PC-centric approach. They failed because PC owners bitched and moaned since their £250 card could not max out Crysis.

Now look back at Crysis nearly a decade later - it looks utterly amazing for such an ancient game.

Its not going to happen anytime soon. Just look at more and more of the PC-centric devs like CDPR?? They moved to consoles too,and look at how much The Witcher 3 got graphically downgraded over initial previews??

It was because of consoles and also probably because CDPR probably realised most people would not have a fast enough graphics card to run the game at those kind of settings anyway.

I think you're either overestimating consoles or underestimating PCs, at least a bit. DX9 lasted long because of Win7 just as much as consoles. DX10 was mostly dead in the water (there were what, like two games developed for it?) and DX11 had very slow adoption rates. DX12/mantle is accelerating quickly though, and that ain't just because of consoles. Yes, a lot of development cycles are spent in the console world, I'm not refuting that, but a lot of it is either now in the PC realm or it's cross-platform.

Putting multi-thread development on consoles is just silly, consoles and PCs have supported MT for what, like a decade now? Many devs just haven't coded for it for various reasons. Plenty of studios release games specifically for PCs without giving a flip about consoles, but outside of that many games are inherently multi-platform due to consoles being basically PCs now. Hardware peculiarities notwithstanding, you had 'modern OSs' running on consoles back in the PS4 era. This isn't a 'warcraft 1 vs gamecube cartridge' world anymore.

Initial previews of games are always to be ignored, because they're almost always 'what we're aiming for' renderings. This has been the case for as long as advertising renders were a thing. The fact remains you can still crank up any console -> PC cross platforms far higher on PCs than on consoles, and if a dev is going to focus on this capability (instead of pushing out a trash console port... see Dishonored 2), they're going to pay attention to what cards are playing those games.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
So when Nvidia is using JPR,its AMD PR helping them?? I know AMD PR has been accused of many things,but that is a new one for me! :p

Nvidia has been using JPR since 2005:

https://www.nvidia.com/object/IO_18301.html

1. I didn't say there's anything wrong with using JPR. Or hiring them for their marketing assistance.
2. You're link is to a press release that quotes JPR. It does not contain any evidence that the report was commissioned by Nvidia.
 

tential

Diamond Member
May 13, 2008
7,355
642
121
The obvious difference between JPR and Steam surveys, is JPR is telling you what is being sold in a particular small time frame. Steam shows what gamers are currently using, which may span many years worth of ownership and doesn't track GPU's being used for other purposes.

I really doubt the opt in and opt out aspect of Steam plays a significant role in results. It just shrinks the data a little.

Whether you doubt it or not doesn't change statistical premises that opt in surveys are not as accurate as true random sampling....

It's just a butchering of statistics. No one is disputing that Nvidia has the lead in marketshare and gpus in use.
It's just not useful to use steam as anymore than just a general guideline when we know it has multiple issues.

You can use tons of more accurate methods to show Nvidia is leading the gpu market. You don't need to dip into steam and the less than accurate data there....