JPRFor the 4th quarter of 2015, GPU shipments increased 2.4% from last quarter

[Mahigan]

Yup, so while Malihagan's comments may seem self centered, I do believe that things ripple through the community. We tech enthusiasts, our conversation trickles down to the general gamer.

And yes, crazy how AMD basically launched the same card twice with WILDLY different acceptance rates from the general community as a whole.

Nvidia took some body blows this cycle, and contrary to everyone's popular belief, including my own, Nvidia isn't invulnerable. If AMD starts to make smart decisions, people will buy their products.

So I don't feel so unoptimistic about the GPU market as I used to. if AMD delivers, we'll continue with multiple competitors. If AMD fails with Polaris, I really don't know how much longer they can actually sell GPUs in worthwhile quantities.

That's how it works. I wasn't trying to be self centered as I wasn't alone when it came to the async controversy. I just didn't want to name drop but the person you're replying too is another individual I'd add to that group.

There are a few more, like the folks at Beyond3D for example.

For me, I find the feature fascinating. So fascinating that I've taken a lot of time to understand the ins and outs of how it works. When the first game released with support for it, obviously I was enthused. Of course, I do have Aspergers so that further compounds the issue for me. If something fascinates me, I want to talk about it all the time.

 

[ShintaiDK]

Isn't that "GPU revenue growth" includes mobile GPUs too? No competition from AMD for that market until polaris.

This thread specifically talked about Desktop GPU shipments, you should read the link the OP posted.

If you look at Steam numbers, its quite clear who won the desktop game. 300 series and Fury series nowhere to be found, GTX960 and GTX970 as top 2 cards.

Even 980 TI got more sales than the entire R7 200 series.

 

[AtenRa]

If you look at Steam numbers, its quite clear who won the desktop game. 300 series and Fury series nowhere to be found, GTX960 and GTX970 as top 2 cards.

Even 980 TI got more sales than the entire R7 200 series.

GPU sales, Current GPUs used and Steam numbers are not the same.

 

[ShintaiDK]

GPU sales, Current GPUs used and Steam numbers are not the same.

Since we can see models, the numbers are quite relevant. And they reflect the current sold products. 300 series and Fury series have been a complete disaster. Not even able to hit the 0.16% to get on Steam. While a card such as the 980TI alone sits at 0.66%.

Its also quite clear from AMDs revenue point of view. And more obvious, from Nvidias revenue.

The card you championed to no end wasn't even to be found in the forum.

 

[AtenRa]

Since we can see models, the numbers are quite relevant. And they reflect the current sold products.

I know you like current STEAM numbers but those are not reflecting global GPU sales.

Also, STEAM hardware survey is not using 100% of steam users, I havent participated in the survey for more than 5-6 months.

300 series and Fury series have been a complete disaster. Not even able to hit the 0.16% to get on Steam. While a card such as the 980TI alone sits at 0.66%.

Have you seen the DX-11 GPUs on STEAM ???
No NVIDIA Card present

There is a huge problem with STEAM statistics the last 3-4 months.

Where are the statistics for GTX 980Ti before January 2016 ???

 

[airfathaaaaa]

so if they are a complete disaster where do that +6% for amd came from? firepro's? x1800xt?

 

[ShintaiDK]

so if they are a complete disaster where do that +6% for amd came from? firepro's? x1800xt?

We have to see in Q1 if it was nothing but inventory dumping as they have done before.

 

[airfathaaaaa]

We have to see in Q1 if it was nothing but inventory dumping as they have done before.

you do understand that this is profit right? and a huge one considering how bad was 2015 for them

 

[ShintaiDK]

you do understand that this is profit right? and a huge one considering how bad was 2015 for them

When you see the revenue difference between NVidia and AMD for Q4 its quite obvious what happened.

 

[airfathaaaaa]

When you see the revenue difference between NVidia and AMD for Q4 its quite obvious what happened.

please do point out WHERE it says about the revenue from each department

 

[tential]

The problem with using steam ShintaiDK in this thread is we have it confirmed amd discrete shipment are up and nvidia are down.

Steam doesn't show that increase or decrease thus we realize that an optional survey gives an idea but doesn't tell the full story. Unless you want to dispute the article ShintaiDK but it's not reflecting the steam numbers youre quoting

 

[ShintaiDK]

If you stuff the channel they are shipped. But shipped isn't the same as sold. Steam is an indicator of actual used units. Its quite clear when higher end cards isn't showing up. Unless the idea is that AMD users dont play games on Steam.

And its nothing new, it happened before.

When you see NVidia with a record quarter, you know why. Where is the money?

This is why YoY is a much better indicator, since it will remove inventory dumping.

And lastly:
http://www.digitimes.com/news/a20160223PD208.html

 

[provost]

For what its worth, I have never opted in for Steam survey through my ownership of any Nvidia cards 690, Titan 780 Ti , or AMD cards for that matter. Nor have a I ever looked at it, before making a purchase decision. So, this may be one data point to gauge adoption of a one vendor's sku over another, but I certainly hope that the various participants (suppliers, developers, etc) linked to the consumer graphic segment are using more sophisticated approaches to plan their resources and project out their forward looking internal budgets.

 

[Abwx]

If you stuff the channel they are shipped. But shipped isn't the same as sold. Steam is an indicator of actual used units. Its quite clear when higher end cards isn't showing up. Unless the idea is that AMD users dont play games on Steam.

And its nothing new, it happened before.

When you see NVidia with a record quarter, you know why. Where is the money?

This is why YoY is a much better indicator, since it will remove inventory dumping.

And lastly:
http://www.digitimes.com/news/a20160223PD208.html

Dont waste your time with unsubstancied data, to know the marketshare all we need is in the OP :

Desktop graphics add-in boards (AIBs) that use discrete GPUs decreased -4.87% from last quarter.

Nvidia’s desktop discrete GPU shipments were down -7.56% from last quarter

AMD’s shipments of desktop discrete GPUs increased 6.69% from last quarter

So what s your take..?.

I say that those numbers point to 56.5/43.5% marketshares in DTGPUs..

 

[Qwertilot]

The steam survey is rather a good data point really - its a healthily sized sample of the PC gaming community and there's surely no reason to suppose any sort of systemic bias?
(Only the installed user base of course but whatever.).

It definitely wouldn't be a big shock - and I couldn't imagine it worrying NV one bit either - if desktop maxwell was starting to tail off. The 970 has been out for 17 months!

 

[USER8000]

I honestly don't see why people are still bickering over simple facts.

AMD gained desktop discrete marketshare from Nvidia.

Nvidia gained laptop discret marketshare from AMD and Intel,but mostly from Intel as AMD had a much smaller dip and it also means that more laptops with Intel chips have an Nvidia GPU in it.

Slightly more PCs now have a discrete GPU in them overall but the PC market has shrunk overall by 10% compared to the previous year,but discrete cards sales dropped at a lower rate by 7% it seems.

Not,rocket science is it?

The steam survey is rather a good data point really - its a healthily sized sample of the PC gaming community and there's surely no reason to suppose any sort of systemic bias?
(Only the installed user base of course but whatever.).

It definitely wouldn't be a big shock - and I couldn't imagine it worrying NV one bit either - if desktop maxwell was starting to tail off. The 970 has been out for 17 months!

Its the only one we have but is biased by the fact it is going to be answered more by hardware enthusiasts than gamers and even in my case,it never asked me when I had my previous Kepler based card,but asked me when I had a GTX960 and most of my mates never answered it,only the hardware enthusiasts like me!

So basically Steam thinks I had an HD5850,and jumped to a GTX960,which is incorrect.

It also meant it never saw I had a Sandy Bridge Core i5 for years either and only saw my Ivy Bridge Core i7.

Plus it probably thought I was using Windows 7 and jumped to Windows 10 too. I didn't - I was using Windows 8 and Windows 8.1 before going to Windows 10.

 

[railven]

I honestly don't see why people are still bickering over simple facts.

AMD gained desktop discrete marketshare from Nvidia.

Nvidia gained laptop discret marketshare from AMD and Intel,but mostly from Intel as AMD had a much smaller dip and it also means that more laptops with Intel chips have an Nvidia GPU in it.

Slightly more PCs now have a discrete GPU in them overall but the PC market has shrunk overall by 10% compared to the previous year,but discrete cards sales dropped at a lower rate by 7% it seems.

Not,rocket science is it?

Is shipping what constitutes market share? I asked this earlier, didn't get a reply. Shipping a product doesn't translate to it being sold.

I used the Xbone example. MSFT ramped up shipping, putting almost 1.1million units out per [EDIT: it was per quarter, not month] versus Sony's PS4 only to have to cease shipping to the US due to excess channel inventory. They redirected units globally where they still got crushed by PS4.

So, is AMD marketshare going up solely based on shipped units? I'm just curious how it is argued for in this forum (seems every forum I frequent has a different interpretation haha. "They shipped more!" "Yeah but they still lost money!" "Don't matter.")

 

[USER8000]

Is shipping what constitutes market share? I asked this earlier, didn't get a reply. Shipping a product doesn't translate to it being sold.

I used the Xbone example. MSFT ramped up shipping, putting almost 1.1million units out per month versus Sony's PS4 only to have to cease shipping to the US due to excess channel inventory. They redirected units globally where they still got crushed by PS4.

So, is AMD marketshare going up solely based on shipped units? I'm just curious how it is argued for in this forum (seems every forum I frequent has a different interpretation haha. "They shipped more!" "Yeah but they still lost money!" "Don't matter.")

The report is talking about marketshare of shipped boards. Even all those reports of Nvidia/AMD having 77%/23% are the same metrics of shipping and its been the case since these have been published even a decade ago.

Generally speaking AMD/ATI has been around 35% of all boards shipped,with brief blimps to 50% during the X800 days usually around 35% or thereabouts.

Even the ENTIRE marketshare estimates can only be based on total shipped boards over a time period.

The problem is all the marketshare reports are based on shipping quantities from AIO partners,so ultimately its why we are looking at percentage changes between quarters,

But,think about it this way - AIO partners would be only shipping more cards if there was channel demand for the cards in prebuilt computer and DIY shops,so it is connected to actual demand for said cards.

So they are linked.

The Steam Hardware Survey OTH is biased by the fact it is optional,does not consistently asked the same people yearly and probably will only mean people who care will answer. It also does not include all the graphics cards used for non-gaming purposes too.

If it was made compulsory it would be a far better metric of seeing what gamers actually use,but again would not be an indication of the total graphics market.

I would suspect quite a fair number of discrete cards are used for non-gaming purposes IMHO OFC.

The only realistic way to get proper metrics would be for Microsoft to data mine all the Windows PCs to see who has what graphics adaptors installed.

Edit to Post.

But the shipped boards per quarter metric JPR uses is far more relevant to investors than total marketshare metrics anyway. The latter would be more important for software companies like Microsoft.

You could argue about actual sales metrics which be ideal,but again even with consoles the metrics will still be total units shipped.

You would need to data mine every retailer in the world which sold a graphics card,PC or console to get those figures. It would be a lot of time and effort and money to do so.

Remember even Microsoft data mining would not cover around 10% to 15% of the PC market.

 

[AtenRa]

Is shipping what constitutes market share? I asked this earlier, didn't get a reply. Shipping a product doesn't translate to it being sold.

I used the Xbone example. MSFT ramped up shipping, putting almost 1.1million units out per [EDIT: it was per quarter, not month] versus Sony's PS4 only to have to cease shipping to the US due to excess channel inventory. They redirected units globally where they still got crushed by PS4.

So, is AMD marketshare going up solely based on shipped units? I'm just curious how it is argued for in this forum (seems every forum I frequent has a different interpretation haha. "They shipped more!" "Yeah but they still lost money!" "Don't matter.")

JPR research in the OP is GPUs shipped, meaning what AMD, Intel and NVIDIA sold themselves to the AIBs and OEMs/ODMs the quarter in question. This is not what people around the world bought for their systems.

Actual GPUs in use around the globe is not even possible to be measured.

And one more thing about Steam Hardware Survey, many people have more than one system this days. They could take the survey this month with the APU system and the next month with the GTX980Ti. Or one month could be with the Notebook and the next with the Fury.

Also is you see what GAMEs each country plays the most, and what is the top 5 dGPUs you will scratch your head.

Example, USA
http://steamspy.com/country/US

Most played games are

1) CS : Global
2) Garry's Mode
3) TF2
4) DOTA 2
5) Rocket League

According to Hardware Survey of January 2016 , GTX970 is the Top dGPU. Does people buy the GTX970 to play those 5 games ??

You will have to go to number 6 and Fallout 4 to really need a card like GTX 970.

That means that most registered Steam Users dont participate in the Hardware Survey each month.

 

[airfathaaaaa]

If you stuff the channel they are shipped. But shipped isn't the same as sold. Steam is an indicator of actual used units. Its quite clear when higher end cards isn't showing up. Unless the idea is that AMD users dont play games on Steam.

And its nothing new, it happened before.

When you see NVidia with a record quarter, you know why. Where is the money?

This is why YoY is a much better indicator, since it will remove inventory dumping.

And lastly:
http://www.digitimes.com/news/a20160223PD208.html

so you claim that because a card isnt showing up on steam survey then it must be that it doesnt sold ?steam survey doesnt have ANY of the 3xx series card on it how can you even say something like this lol its a well known bug since 2014 that they customers support said that it will get fixed(valve time ofc...)

also nvidia had a record quarter because of the notebook sales unless ofc you mean that -7 on the gpu department somehow made them earn more in a magical way
also invetory dumping happens when they do have attractive sales and since the only thing they did was to cut the price a bit of nano and fury air cooled i doubt that is the case for a YoY as you say in order of an dumping to happen

you just simply dont want to accept that people are buying amd cards not because they are huge bairgain but because they are aging way better than cards that last a year or two

 

[tential]

If you stuff the channel they are shipped. But shipped isn't the same as sold. Steam is an indicator of actual used units. Its quite clear when higher end cards isn't showing up. Unless the idea is that AMD users dont play games on Steam.

And its nothing new, it happened before.

When you see NVidia with a record quarter, you know why. Where is the money?

This is why YoY is a much better indicator, since it will remove inventory dumping.

And lastly:
http://www.digitimes.com/news/a20160223PD208.html

Good point you're right my mistake shipped is not sold.

I misread tje op and thought it was sold gpus. In that case ShintaiDK is absolutely right steam is a better indicator. If we saw improvement in amd numbers whatsoever we could conclude amd is gaining traction. Really all we can conclude is that vendors are asking for more stock of amd products.

 

[USER8000]

Good point you're right my mistake shipped is not sold.

I misread tje op and thought it was sold gpus. In that case ShintaiDK is absolutely right steam is a better indicator. If we saw improvement in amd numbers whatsoever we could conclude amd is gaining traction. Really all we can conclude is that vendors are asking for more stock of amd products.

Steam wouldn't be a better indicator. Firstly around 10% to 15% of all PCs don't run Windows.

Then take the Windows PCs,and what percentage of systems with dGPUs will be used for gaming under Windows??

Then look at some games like WoW,which don't use Steam.

Then,add the fact the Steam survey is optional,and people like me had a break of a few YEARS to be asked,when I changed most parts of my system.

Unless you really honestly think more GTX970 cards were sold than GTX750 or GTX750TI derived cards in ALL PCs and all GTX960 derived cards in ALL PCs.

You might want to see what many prebuilt PCs and laptops are using.

Loads of the ones under £1000 even in the UK or Germany are mostly using GM107 and GM206 based chips.

The GTX750 and GTX750TI were out 10 months before the GTX970 alone. Nvidia and AIB partners have consistently said its a very popular card to the extent that they have been even using GM206 chips to meet the demand.

Its even been proven before on this very forum,the Steam metrics are quite wonky in picking up marketshare changes.

Its a useful data point but in the end it is deeply flawed too.

I would rather trust units shipped over an optional survey which only a small percentage of actual gamers will answer.

Valve has not even told us the exact sample number too.

I still expect lots of dGPUs to be sold for non-gaming use on top of this.

Even the units ship method is flawed,so in the end we can only get some rough indications of where changes are happening.

Only Microsoft will have any decent reliable metrics of actual Windows PC dGPU marketshare.

 

[Hitman928]

IMO, discussing steam hardware survey right now is moot, it seems to be broken anyway. I don't know if it's the transition to Windows 10 or what, but it's definitely having issues.

Look at that resurgence of DX8 GPUs away from DX11. Who knew they'd be so popular in 2016? What goes around comes around I guess. They're also missing two of the last five months of data. Lots of weird things in there, I just listed two of the big ones.

 

[tential]

It's a positive indicator, we need AMD sales numbers to verify really.

 

[maddie]

Good point you're right my mistake shipped is not sold.

I misread tje op and thought it was sold gpus. In that case ShintaiDK is absolutely right steam is a better indicator. If we saw improvement in amd numbers whatsoever we could conclude amd is gaining traction. Really all we can conclude is that vendors are asking for more stock of amd products.

I guess that these vendors must be hoping to be 1st in line for Polaris. Imagine ordering GPU dies that never get sold. Polaris must be fantastic.

Just to be clear to everyone, I'm being partially facetious.

Is it really being implied that GPUs are ordered from AMD but never get shipped or sold. Oh my.