The balance of power in gaming (Jon Peddie Research)

cageymaru

cageymaru

Fully [H]
Joined
Apr 10, 2003
Messages
21,723
The balance of power in gaming.
http://gfxspeak.com/2016/11/16/balance-power-gaming/

57% of gamers are running on AMD Radeon graphics.
http://www.pcgamesn.com/amd/57-per-cent-gamers-on-radeon


The gaming hardware market is huge, and bewildering because of its diversity and the marketing hype associated with it. We’ve been monitoring and measuring the gaming market for over a decade, and felt it was time to share some of our findings and clear up some of the misunderstandings about the market and the suppliers.

The major take away is when x86-based gaming is examined, AMD is the clear winner.

Gamers fall into several camps, device size, main processor type, OS, screen size, budget, time spent, etc. The gaming community is so large; no single demographic can define it.


Guess that's why their stock price is over $8.51 a share right now.

Discuss.

 
The stock price is what it is due to the Nvidia blowout, its called robot trading. And its hard to take those links serious :)

AIB numbers should soon be out.

https://www.jonpeddie.com/press-rel...-q3-2016.-total-gpu-shipments-up-a-whopping-2

MW_table_q316.PNG


MWq3163_580-2.png
 
Is this what we're gonna do now. Lump consoles and APUs into the figure to fluff the bad news?

I bought an AMD laptop for my 63 year old mother a few months ago... She is not a gamer. And you can thank Trump for the stock boost, pretty sure everyone is breaking records.
 
I can only say this is an interesting way to produce numbers.

Any target audience meetings I attend, the numbers are never broken down this way. It's always broken down into PS3, PS4, Xbox 360, Xbox One, Wii/Wii U, 3DS, Windows, Mac OS, Linux. That's basically how many builds we have to produce to target these individual markets.

We have yet to do a Windows AMD or Windows Nvidia build....


Edit: now that I scrutinized the numbers more. They don't jive with Steam's Hardware Survey in terms of their marketshare?
 
ChrisC said:
Edit: now that I scrutinized the numbers more. They don't jive with Steam's Hardware Survey in terms of their marketshare?
Click to expand...
Steam HW Survey is notoriously inaccurate (particularly under-cutting AMD GPUs).
Total marketshare includes miners and non-gamers, two categories that heavily favor AMD / Intel.
 
ChrisC said:
Edit: now that I scrutinized the numbers more. They don't jive with Steam's Hardware Survey in terms of their marketshare?
Click to expand...


The numbers really won't jive with the steam survey numbers since JPR numbers are a time slice and Steam's numbers are progressive and only talk about people playing games on their servers.
 
cageymaru said:
The balance of power in gaming.
http://gfxspeak.com/2016/11/16/balance-power-gaming/

57% of gamers are running on AMD Radeon graphics.
http://www.pcgamesn.com/amd/57-per-cent-gamers-on-radeon


The gaming hardware market is huge, and bewildering because of its diversity and the marketing hype associated with it. We’ve been monitoring and measuring the gaming market for over a decade, and felt it was time to share some of our findings and clear up some of the misunderstandings about the market and the suppliers.

The major take away is when x86-based gaming is examined, AMD is the clear winner.

Gamers fall into several camps, device size, main processor type, OS, screen size, budget, time spent, etc. The gaming community is so large; no single demographic can define it.


Guess that's why their stock price is over $8.51 a share right now.

Discuss.
Click to expand...

Isn't their stock price more due to a contract google rather than controlling 57% of AMD Radeon graphics.
 
Chimpee said:
Isn't their stock price more due to a contract google rather than controlling 57% of AMD Radeon graphics.
Click to expand...

Nah, the google contract is quite small. Its not even for HPC or AI. Traders just see GPU as one. So if it goes well for Nvidia, it must go well for AMD. Add automated trading on top for the full effect. Even AMD couldn't tell you why this happens. They just sold stocks at 6$.

Funny thing. Using the same math as used here, you end up with something like more people than on the planet plays on ARM devices.
 
Shintai said:
Nah, the google contract is quite small. Its not even for HPC or AI. Traders just see GPU as one. So if it goes well for Nvidia, it must go well for AMD. Add automated trading on top for the full effect.

Funny thing. Using the same math as used here, you end up with something like more people than on the planet plays on ARM devices.
Click to expand...

I see, but I thought the google contract was directly the cause of why AMD gone to past 8 bucks from 7 bucks in their stocks.
 
Chimpee said:
I see, but I thought the google contract was directly the cause of why AMD gone to past 8 bucks from 7 bucks in their stocks.
Click to expand...
Chimpee said:
Isn't their stock price more due to a contract google rather than controlling 57% of AMD Radeon graphics.
Click to expand...

Yea, their price jumped on Google's decision to go AMD not because of some research paper. And really, these financial, market saturation, and stock earnings reports are just fodder for biased soap box posting. Give it time and the same 4 characters will be all over this.
 
The one bit of good news AMD gets and we hear...


1. Power consumption sucks
2. No R&D dollars
3. Steam surveys are what matters
4. AMD drivers suck
5. Polaris sucks
6. Roy Taylor trolling
7. Zen will be slow and use 220w and have slower than BD single threaded performance.
8. What else am I missing to make AMD look bad?
9. Go away trolls
10. Thank you
 
Rvenger said:
The one bit of good news AMD gets and we hear...


1. Power consumption sucks
2. No R&D dollars
3. Steam surveys are what matters
4. AMD drivers suck
5. Polaris sucks
6. Roy Taylor trolling
7. Zen will be slow and use 220w and have slower than BD single threaded performance.
8. What else am I missing to make AMD look bad?
9. Go away trolls
10. Thank you
Click to expand...

No shit. I mean spot on no shit. Anything to belittle AMD and good news. NV had one of its best qtr's ever and congrat's to them but anything positive AMD is crapped on and or not possible, or just paper shuffling. You can tell which website your on.
 
TaintedSquirrel said:
Is this what we're gonna do now. Lump consoles and APUs into the figure to fluff the bad news?

I bought an AMD laptop for my 63 year old mother a few months ago... She is not a gamer. And you can thank Trump for the stock boost, pretty sure everyone is breaking records.
Click to expand...

Stop spoiling the fun dammit.

Wait, what was the % profit and high-end share of AMD again?
 
razor1 said:
The numbers really won't jive with the steam survey numbers since JPR numbers are a time slice and Steam's numbers are progressive and only talk about people playing games on their servers.
Click to expand...

Something that seems to be lost is that Steam's numbers are direct survey numbers with end users (actually bought and being used) whereas the JPR reports (and other similar) have always been shipment estimates from the IHVs (this means products sitting at distributors, retailers or even waiting to be integrated by OEM/AiBs are all included). It's worth noting that AMD has for quite awhile now been more aggressive in terms of actual shipments compared to demand relative to Nvidia who by contrast more tightly controls supply (hence also why you tend to run into the impression that Nvidia's is more expensive rising above MSRP whereas AMD is cheaper often below MSRP).

It's also worth noting, for those that don't like Steam's numbers, that this specific report actually derives (manipulates, so possibly more inaccuracy due to error/bias) its estimates using Steam's numbers. So if you question accuracy and relevance of Steam's numbers, what about numbers derived from them (this report)?

Based on Steam user data, we conclude that 16.5% of desktop machines used for gaming are using Intel integrated graphics (Southeast Asia represents about 6.5 percent of Valve’s global Steam gamers). Steam data does not differentiate desktop from notebook integrated graphics. Therefore, we applied the same percentage to Intel for notebooks and based on the AMD to Intel integrated notebook ratio (our quarterly data collection of mobile integrated GPU shipments), normalized AMD’s notebook integrated gamer utilization.

We also used steam data to estimate the market share for AMD and Nvidia mobile discrete GPUs used for gaming based on our quarterly data collection of mobile discrete GPU shipments.
Click to expand...
 
Last edited:
Does the Steam Survey list how many systems were polled? I haven't been polled in many months and my PC runs 24/7. Just wondering if the Steam Survey is for 10,000 systems polled, or 100,000,000. Have they started doing mandatory polling?
 
For people questioning sample sizes and selection.

A sample size of 100k, which is less than 1 in 1000 given Steam's user base, would result in a margin of error of something like 1% with 99% probability (even better, this is just off the top of my head but you can calculate this in detail). This would be a level of accuracy already well beyond what the majority of polling data you see used in other topics.

So what Steam has a selection or report bias? So for whom and why? If anything a random selection (which is what would be used to combat bias) naturally would mean people should not be repeatedly polled. If they only repeatedly poll past respondents and/or the same people that would introduce selection bias.

Or respondents, I guess this means users of certain hardware, themselves has a response bias? Again what would possibly be the reasoning here.

Also if we use this argument then why is the data in this JPR report relevant at all? It itself uses Steam data. Also how do people think those industry shipment reports are estimated? That JPR (or whomever) actually goes around and gets hard numbers from what every single retailer? distributor? IHVs themselves? Those are also parties that are even more likely to want to keep data to themselves for privacy reasons than individual Steam users.

Not to mention a lot of the counter arguments on why people feel the data is not representative is even more poor and completely anecdotal.
 
None of these threads are relevant because they all use the same flawed data. They are a means to shill.
 
Total GPUs sold is what matters. Just add them up and that tells what your installed base is. Of course stores don't like to release their sales numbers and thus you will never get GPU manufacturers to do the same. With that said, I think it's nice to look at the market from the eyes of a stock market analyst. A trader on Wall Street is going to base their stock buying habits on someone's research no matter how flawed or truthful it is.

In the end it's what the people with influence and investors think that matters. They dictate who is the current darling and belle of the ball. Whomever smells more like roses gets the influx of capital.


Here is another article for you'll to discuss their contextualization of the research.

Over 400 Million Gamers Are Using Radeon Graphics in New Study
http://www.legitreviews.com/400-million-gamers-using-radeon-graphics-new-study_188127

Guess what study it references? :)

Bet more people are likely to agree with this one from the SAME firm.
http://jonpeddie.com/publications/market_watch/

Same folks; spinning information in various ways for the their varying target audiences.
 
cageymaru said:
Total GPUs sold is what matters.
Click to expand...

Nope, only end profits matter, because that is entire point of running a business. Even if we take the misleadingly bloated AMD numbers at face value, all it does is proving how terrible AMD is at extracting money off their huge marketshare.

Samsung will kill their own mothers to switch places with Apple in smartphone profits, but we still have dumb "analysts" fixating on how bad Apple marketshare since forever as if it that is the only thing that matters.
 
Palladium@SG said:
Nope, only end profits matter, because that is entire point of running a business. Even if we take the misleadingly bloated AMD numbers at face value, all it does is proving how terrible AMD is at extracting money off their huge marketshare.

Samsung will kill their own mothers to switch places with Apple in smartphone profits, but we still have dumb "analysts" fixating on how bad Apple marketshare since forever as if it that is the only thing that matters.
Click to expand...

Even better. I read someone say that because the consoles don't advertise Radeon on the splash screen that those sales don't matter. You twisted it perfectly to make it sound good from an Nvidia or Intel standpoint. It's all in how you interpret the data.

I was looking at the Steam Survey data and they had an exhaustive list of Nvidia DX12 video cards. I looked at Civilization 6 and their DX12 implementation only works on Maxwell or better. I could drop Nvidia's real world DX12 market share by spinning the numbers a bit to reflect that a great deal of their GPUs aren't DX12 compliant with real world games.

It's all in how you look at the numbers and how intelligently you can spin it to suit your needs.
 
cageymaru said:
Does the Steam Survey list how many systems were polled? I haven't been polled in many months and my PC runs 24/7. Just wondering if the Steam Survey is for 10,000 systems polled, or 100,000,000. Have they started doing mandatory polling?
Click to expand...

They most likely poll over a million. However they have shown month after month that they are consistent. So the poll is as such completely accurate.
 
cageymaru said:
Over 400 Million Gamers Are Using Radeon Graphics in New Study
Click to expand...
And only 100 million of those (est.) even heard of word "Radeon" :)
cageymaru said:
I could drop Nvidia's real world DX12 market share by spinning the numbers a bit to reflect that a great deal of their GPUs aren't DX12 compliant with real world games.
Click to expand...
Don't forget to drop GCN1.0 in the process. Spinning is fine as long as it is consistent.

Anyways, for all that matters. If AMD dominates GPU shipment market. Why were they stuck in the hellhole again?
 
Well adding consoles into the mix AMD definitely ships more GPU's than another company that is a given. But Console market doesn't really influence the PC market (if the opposing GPU makers can make hardware that can run those console games on PC's just as well as the Console manufacture's GPU's) and that should be a resounding conclusion from these new numbers.

So to answer
lolfail9001 said:
And only 100 million of those (est.) even heard of word "Radeon" :)

Don't forget to drop GCN1.0 in the process. Spinning is fine as long as it is consistent.

Anyways, for all that matters. If AMD dominates GPU shipment market. Why were they stuck in the hellhole again?
Click to expand...

Yes you are right most console users don't know anything about what hardware is inside their consoles. Most console users just know how to turn their consoles on, install a game and play, they don't know what software that game runs on, they don't know the hardware, they don't know anything that a person that visits a hardware tech forum would know, even the basics. These are the same type of people that don't know the difference between Android, Apple, Microsoft phones, outside of what they "feel" because others have told them somethings.

Why they are stuck in the hellhole because console sales don't garner much for AMD when it comes to margins or profits. Which if we follow Q after Q financial calls it should be understood.

So anyone trying to use these numbers to show "how well" AMD is doing in the GPU marketing just doesn't know he/she is talking about. Its not about numbers, but what those numbers represent as an end result for the each company is what matter.
 
cageymaru said:
Does the Steam Survey list how many systems were polled? I haven't been polled in many months and my PC runs 24/7. Just wondering if the Steam Survey is for 10,000 systems polled, or 100,000,000. Have they started doing mandatory polling?
Click to expand...
What I've found interesting I get the Steam survey repeatedly on my Nvidia GPU systems and for the life of me can't remember when on my AMD GPU systems. My laptop the other day (860m), Last week on the 1060 Node 202 rig and once again on my 1070 rig and zero for the Nano or 290 rig. WTF is that just chance or what??? From my experience Steam survey is shit!
 
razor1 said:
Well adding consoles into the mix AMD definitely ships more GPU's than another company that is a given. But Console market doesn't really influence the PC market (if the opposing GPU makers can make hardware that can run those console games on PC's just as well as the Console manufacture's GPU's) and that should be a resounding conclusion from these new numbers.

So to answer


Yes you are right most console users don't know anything about what hardware is inside their consoles. Most console users just know how to turn their consoles on, install a game and play, they don't know what software that game runs on, they don't know the hardware, they don't know anything that a person that visits a hardware tech forum would know, even the basics. These are the same type of people that don't know the difference between Android, Apple, Microsoft phones, outside of what they "feel" because others have told them somethings.

Why they are stuck in the hellhole because console sales don't garner much for AMD when it comes to margins or profits. Which if we follow Q after Q financial calls it should be understood.

So anyone trying to use these numbers to show "how well" AMD is doing in the GPU marketing just doesn't know he/she is talking about. Its not about numbers, but what those numbers represent as an end result for the each company is what matter.
Click to expand...
I would think if you use APU's from AMD which I think is valid plus x86/a64 then consoles should be included. Then again Nvidia also has APU's, maybe not x86/a64 but they will be used in next Nintendo console plus other applications as well. So looks like GPU wise they are about equal. Wait! Where is Intel on this and their APU's :ROFLMAO: - OOPS Intel blows both of them away even combined together!
 
If you take AMD as a company, instead of just the GPU market IF Zen does well and does take off, AMD won't need to worry about Nvidia.

Of course then you wake up and look at reality and realize AMD will probably blow the Zen launch as well.
 
Brackle said:
If you take AMD as a company, instead of just the GPU market IF Zen does well and does take off, AMD won't need to worry about Nvidia.

Of course then you wake up and look at reality and realize AMD will probably blow the Zen launch as well.
Click to expand...


Oh no no no, AMD still has to worry about nV, they still have to worry about Intel, even if Zen is good, that doesn't' mean they will automatically gain marketshare even if its a good launch. The only thing it looks like Zen will do in the short term is improve margins, for them to to start expanding with what they have right now it will take some time.
 
noko said:
What I've found interesting I get the Steam survey repeatedly on my Nvidia GPU systems and for the life of me can't remember when on my AMD GPU systems. My laptop the other day (860m), Last week on the 1060 Node 202 rig and once again on my 1070 rig and zero for the Nano or 290 rig. WTF is that just chance or what??? From my experience Steam survey is shit!
Click to expand...


Well I haven't gotten a single steam survey for the past 3 years lol.
 
razor1 said:
Well I haven't gotten a single steam survey for the past 3 years lol.
Click to expand...

Statistically you will get a steam survey poll once every 13 years or less. So its always fun when someone complains that THEY never got polled.

I tried it a single time in my entire steam enabled life. And it wasn't even my account but my better half's account on her PC.

But again, we know the data is correct month after month, year after year.
 
Shintai said:
Statistically you will get a steam survey poll once every 13 years or less. So its always fun when someone complains that THEY never got polled.

I tried it a single time in my entire steam enabled life. And it wasn't even my account but my better half's account on her PC.

But again, we know the data is correct month after month, year after year.
Click to expand...
Then WTF!? It happen each time after Steam updated, now I didn't try with the 290 lately but will have to see now. Now I only replied on the 1060 and not on the other two machines, I wonder if that made a difference.
 
AzixTGO said:
if the rumors are true, zen is going to take market share. Nobody is passing up $1000 CPU performance for $300. I would ditch my i7 so fast...

I would think most realize AMD has the most GPUs between them and nvidia. Numbers always showed that. Discrete was the issue.
Click to expand...


Rumors don't tell us how market share changes, product lines and what they are capable of and with the addition of how the market is right now do, Coming out with something that has already been out for 3 or more years is where AMD Zen rumors are at right now and if that is the case if you believe the rumors that means market share won't change, at least not in the short term with these types of markets. The only place I can see Zen pick up any decent amount of marketshare in the short term is mobile. Desktop's won't be out till 6 months down the road and server versions just won't cut it cause Intel has chips that are just as capable, if Intel was dumb enough not to initiate or match a price war with AMD (what ever way it happens) it will be their inability to foresee how the market will react to Zen. This has nothing to do with Zen though. And to your point on price differential, I do agree if Intel doesn't do what I just mentioned its all on Intel, but is that something we should expect? I mean would it even be rational to expect Intel not see that coming and for them to respond, its not like Zen is some mysterious product that they can't compete with in price.

Who the hell cares about APU's? Really you want to throw that out there where they make crap from APU's as any type of justification for none existence discrete in some segments and outclassed in some segments?
 
AzixTGO said:
if the rumors are true, zen is going to take market share. Nobody is passing up $1000 CPU performance for $300. I would ditch my i7 so fast...

I would think most realize AMD has the most GPUs between them and nvidia. Numbers always showed that. Discrete was the issue.
Click to expand...

how are you so sure that AMD Zen is going to offer 1000$ Intel Performance at 300$?.. :eek:
 
razor1 said:
Oh no no no, AMD still has to worry about nV, they still have to worry about Intel, even if Zen is good, that doesn't' mean they will automatically gain marketshare even if its a good launch. The only thing it looks like Zen will do in the short term is improve margins, for them to to start expanding with what they have right now it will take some time.
Click to expand...

CPU's give more profit then GPU's do. If Zen does well (...again this is AMD) Then it is possible AMD can have a very good CPU, combine that CPU tech with Vega GPU ability. It is possible AMD will gain even more market share in the GPU scene and CPU scene at the same time.

That is something Nvidia cannot do. Again this is all speculation because AMD just doesn't do launches very well.

But if ZEN is that good, and VEGA is good, imagine a 1024core vega with a 4-8core ZEN APU.....PEople might go that route for gaming if it does work.

I was able to play rust/dota/wow/cs:go on my A10 laptop @ 720p. Again AMD has to basically be perfect for that to work out.
 
You must log in or register to reply here.
Back
Top