The AMD Execution Thread [2007 - 2017]

[CarstenS]

Steam hasn't been properly detecting AMD graphics cards for years, besides:
And a whopping 20% of these performance oriented users are using Intel graphics

Not arguing that Steam is not the most reliable of sources, but:
- Even gamers use Laptops, many to play less demanding games for example while in the wait queue for their Multiplayer game of choice. Others like myself who are blessed with slow internetz at home use it to download nice 50 GB install files for Doom (just to see there's a 12 GB patch waiting as well) to save power. And guess what: I'm explicitly telling my Laptop to use the IGP for stuff like that and not it's Radeon GPU.

- Detecting AMD cards that present themselves via driver only as "HD 7900" or similar is not the sole fault of Steam. Could they do better? Yes, but that isn't the HW survey's main objective. Besides, it's often also detecting only broad classifications on Nvidia and Intel as well - you mentioned yourself "Haswell".

 

[Deleted member 13524]

Others like myself who are blessed with slow internetz at home use it to download nice 50 GB install files for Doom (just to see there's a 12 GB patch waiting as well) to save power. And guess what: I'm explicitly telling my Laptop to use the IGP for stuff like that and not it's Radeon GPU.

You're just adding another one to the thousands of anecdotal evidences that the hardware survey isn't reliable for looking at the GPU market for gamers. Steam will never detect you Radeon GPU, so that's one more Intel GPU and one less Radeon GPU in the statistics.

 

[Kaotik]

Not arguing that Steam is not the most reliable of sources, but:
- Even gamers use Laptops, many to play less demanding games for example while in the wait queue for their Multiplayer game of choice. Others like myself who are blessed with slow internetz at home use it to download nice 50 GB install files for Doom (just to see there's a 12 GB patch waiting as well) to save power. And guess what: I'm explicitly telling my Laptop to use the IGP for stuff like that and not it's Radeon GPU.

- Detecting AMD cards that present themselves via driver only as "HD 7900" or similar is not the sole fault of Steam. Could they do better? Yes, but that isn't the HW survey's main objective. Besides, it's often also detecting only broad classifications on Nvidia and Intel as well - you mentioned yourself "Haswell".

It's not just about detecting as "7900 series" or something like that, there are clear issues with the system itself.
Just look how high for example HD 8800 Series is - that's OEM Only product which was sold for less than a year, there's simply no way it belongs up there higher than HD 7800 series for example (retail & oem, sold for 2+ years)

 

[Deleted member 13524]

It's not just about detecting as "7900 series" or something like that, there are clear issues with the system itself.
Just look how high for example HD 8800 Series is - that's OEM Only product which was sold for less than a year, there's simply no way it belongs up there higher than HD 7800 series for example (retail & oem, sold for 2+ years)

And the HD7700 series (HD7770/50 + HD7790 for 6 months) has almost 2x more presence than the entire R9 200 range (R9 280X, R9 280, R9 290X, R9 290, R9 285).
Somehow the number of "DX8 and below" GPUs grew 600% within 3 months. Boy, there must have been millions of Radeon 8500 and Geforce 3/4 cards going into new systems this summer.


That whole thing is a crapfest and its info is useless compared to sales numbers between two independent consultants that show very close results despite measuring them in a different way.

 

[DavidGraham]

The tool still remains an important indicator for hardware uptake rate and trends, and it shows data we have seen in real life, like the popularity of 970, 960, 750Ti, Intel Graphics ..etc. AMD's irregularities are affected by several trends: like dwindling market and mind share, which causes low sales (eg, Fury line), people migrating from AMD to NV (eg, R9 200 owners buying Maxwell and Pascal cards which causes downspike of it's percentage), non gamers buying large quantities of GPUs (eg, miners buying the R9 200 series). That and several other factors as well, which when taken into account explains a lot of what seems like major defects. Which makes Steam survey very much relevant in the gaming spectrum.

 

[entity279]

I agree that there are good reasons behind most irregularities, and that's about the only point your post succeeds at covering.

However, how do we know that numbers that appear ok are not irregularities themselves? The fact that a good proportion of the numbers indicated by the survey cannot be taken at a face value, (they need to be interpreted considering "That and several other factors") doesn't comfortably prove its relevance. It much rather puts a dent into that..

Edit : sorry, re-read your post. Seems like you are suggesting the Steam numbers should be taken at face value actually. In that case, fine - nothing to further say

 

[Razor1]

Q2 is always the weakest quarter.

Doesn't matter if its weak or not, if you see what the CEO of AMD stated about the numbers she alluded to rx sales and more of them.

Our strong second-quarter graphics performance was capped by the launch of our new Polaris based RX 480 GPUs at the end of June, which helped contribute to our highest desktop channel GPU shipments since the fourth quarter of 2014. The Radeon RX family resets expectations around the experiences and features gamers now want in a mainstream GPU.

This is BS, rx 480 wasn't launched till the June 29th, it only had one day in Q2 sales! And with a weak quarter, they didn't gain marketshare by selling more cards lol.

AMD CEO just lied to their investors and share holders!

Quarterly figures came out after she stated this, there was no way for investors or shareholders to verify what she was saying was true or not.

 

[no-X]

This is BS, rx 480 wasn't launched till the June 29th…

AMD doesn't sell GPUs to end users, but GPUs to AIB partners. AIBs have to buy the product from AMD weeks or months before launch.

 

[Kaotik]

AMD doesn't sell GPUs to end users, but GPUs to AIB partners. AIBs have to buy the product from AMD weeks or months before launch.

not to mention OEMs getting their cards before retail launch to have time to include them in their new machines

 

[Razor1]

AMD doesn't sell GPUs to end users, but GPUs to AIB partners. AIBs have to buy the product from AMD weeks or months before launch.

I know that, but we also know the supply issues with the rx480, its easy to figure all this out, a 20% drop in total AIB sales, is what 2-2.5 million cards dropped? That hit wasn't taken by AMD, because AMD isn't selling as many cards as nV (GPU's + ram packages) how ever you want to put it.

But JPR doesn't do shipment numbers from AMD. Mercury does.

JPR is total AIB sales, total retail sales. (granted retailers still have to sell them, but the margin of error is going to drop as you get down the supply chain)

AMD didn't gain marketshare by selling more to end users, or OEM's for that matter, yes AMD does sell full boards to OEM's, OEM's don't make their own PCB's as AIB's do. They gained marketshare for the most part by the weakness in nV's maxwell 2 demand, because Pascal has been launched and had supply issues.

 

[Razor1]

not to mention OEMs getting their cards before retail launch to have time to include them in their new machines

So lets see Dell just include the rx480 in an advertisement, the alienware system they talked about is still not up with the rx480, June 29th right? Where is it, that's 1 half months already, and 2 weeks since the advertisement.

Now if you look at HP, they have their box type system, build the system out, and the rx480 holds shipment date of the system back 2 months lol, where are they?

Everything points to the fact, that AMD didn't have this card to sell till June 29th, OEM's and system builders didn't get this card till around that time, and they haven't even procured the card yet to put into their systems.
They could have validated the card a month before launch but they aren't going to send AMD money without knowing when they are getting the cards. They will need steady supplies and that is not happening as of yet.

The HP omen box, as of last week had the rx480 in the build, now its not available at all

http://www.digitaltrends.com/computing/hp-omen-cubed-gaming-desktop-laptop-steelseries/

http://store.hp.com/us/en/mdp/towers-349501--1/hp-omen-x-900-gaming-desktop-pc-412003--1#!

So Oct now or possible even never as it might have been dropped altogether? Are they really selling this GPU's as they say, or are they so supply constrained, that they took orders full well they won't be able to send them out?

PCper also got the reporting right as well

There are several other factors to watch with this data however. First, the quarterly drop in graphics card sales was -20% in Q2 when compared to Q1. That is well above the average seasonal Q1-Q2 drop, which JPR claims to be -9.7%. Much of this sell through decrease is likely due to consumers expecting releases of both NVIDIA Pascal GPUs and AMD Polaris GPUs, stalling sales as consumers delay their purchases.

 

[CarstenS]

You're just adding another one to the thousands of anecdotal evidences that the hardware survey isn't reliable for looking at the GPU market for gamers. Steam will never detect you Radeon GPU, so that's one more Intel GPU and one less Radeon GPU in the statistics.

Really, do I now? Of course it won't detect the Radeon GPU in the system where I don't allow it to run. Why should it be detected if it's never gonna be used? Steam also does not detect the dozens of graphics cards in cardboxes in my closet - because they're not being used.

Don't get your point besides much hate for the Steam survey.

 

[Deleted member 13524]

I'm thoroughly enjoying Razor1's mental acrobatics to somehow justify his opinion about how the AMD's CEO is liyng just by mentioning the results of two independent studies from two different consultants showing marketshare growth in two quarters in a row.

And of course that all those people who abstained from buying a 960/970/980 during the whole H1 2016 did it because they were waiting for the Pascal line (because most gamers keep trace of rumored release schedules, right?). It was certainly not because the Fiji/Hawaii/Tonga cards kept showing better performance than their Maxwell 2 equivalents with practically every new DX12 benchmark that came up. Future-proofing is a totally overblown concept.

Of course it won't detect the Radeon GPU in the system where I don't allow it to run. Why should it be detected if it's never gonna be used?

So you don't care that a survey made by the largest gaming digital distribution channel won't even try to figure out which GPU each system uses for gaming.
It's okay if you don't care. What's not okay is to use these botched survey results to somehow try to discredit actual GPU sales numbers.

 

[firstminion]

Doesn't matter if its weak or not, if you see what the CEO of AMD stated about the numbers she alluded to rx sales and more of them.



This is BS, rx 480 wasn't launched till the June 29th, it only had one day in Q2 sales! And with a weak quarter, they didn't gain marketshare by selling more cards lol.

AMD CEO just lied to their investors and share holders!

Quarterly figures came out after she stated this, there was no way for investors or shareholders to verify what she was saying was true or not.

You need to research what shipments means, Razor1.

 

[Razor1]

tell me then cause I can pull out posts Dave B stated over 7 years ago about the differences between Mercury and JPR numbers. I have stated the exact same thing he has from back then just in different wording.

Yeah at that time I didn't understand why there were differences between Mercury and JPR numbers, but point blank he posted, JPR numbers are AIB sold cards, vs Mercury numbers IHV shipment of GPU's, not final sales.

Either he posted something wrong at that time, or something is wrong now. Which ever it is, I'm gathering, AMD's CEO wasn't very truthful, as Q3 numbers if we see an increase in marketshare because of higher volume of nV cards and the % shift goes back to a more reasonable number like 2.5%; actually the shift was 2.5% in Mercury numbers for this quarter, of total shipped, yet JPR AMD's % increased much more, hmm, interesting, so this is mental gymnastics when you don't even need to look at line items in the financials to see what happened?

I hate to ask you guys to do strenuous weight lifting with your minds then......

 

[CarstenS]

So you don't care that a survey made by the largest gaming digital distribution channel won't even try to figure out which GPU each system uses for gaming.
It's okay if you don't care. What's not okay is to use these botched survey results to somehow try to discredit actual GPU sales numbers.

Funny how you make it sound like I wasn't caring or that I was the one doing what you so despise. I actually said quite the opposite of you claim to have understood. Please check your facts as well as your implications. That's just bad, bad style.

 

[firstminion]

Doesn't matter if its weak or not, if you see what the CEO of AMD stated about the numbers she alluded to rx sales and more of them.



This is BS, rx 480 wasn't launched till the June 29th, it only had one day in Q2 sales! And with a weak quarter, they didn't gain marketshare by selling more cards lol.

AMD CEO just lied to their investors and share holders!

Quarterly figures came out after she stated this, there was no way for investors or shareholders to verify what she was saying was true or not.

Our strong second-quarter graphics performance was capped by the launch of our new Polaris based RX 480 GPUs at the end of June, which helped contribute to our highest desktop channel GPU shipments since the fourth quarter of 2014. The Radeon RX family resets expectations around the experiences and features gamers now want in a mainstream GPU.

A second-quarter "hard launch" needs GPU shipments on that same quarter. She can compare those numbers internally (it's her company) and announce them to investors.

What is the basis on your accusation? Corporate wording are very precise.

 

[Dave Baumann]

tell me then cause I can pull out posts Dave B stated over 7 years ago about the differences between Mercury and JPR numbers. I have stated the exact same thing he has from back then just in different wording.

FYI, the wording of JPR's current report exclusively talk about shipments. Follow console forum to know all about the difference between POS and shipments! Don't know when his focus / wording changed to shipments.

http://jonpeddie.com/publications/market_watch/

 

[Razor1]

A second-quarter "hard launch" needs GPU shipments on that same quarter. She can compare those numbers internally (it's her company) and announce them to investors.

What is the basis on your accusation? Corporate wording are very precise.

There internal shipments right?

Mercury numbers should show that?

Do you see that here?

Highest desktop channel since 2014 , look at desktop discrete, 22.8% that is lower than 2015 q2!
Now with the drop in overall shipments you think, AMD's shipments have increased?

AMD’s average total share increased for the first time in more than four years. The last time the company experienced an increase was the first quarter of 2012. Both desktop and mobile shares were up this quarter

Overall unit shipments of PC graphics fell by 6.5 percent in the second quarter of 2016, and were down 3.1 percent compared to a year ago. Due to a strong shift towards high-end graphics driven both by new products launched in the quarter and fewer low-end units shipped, average
desktop selling prices surged and set a new record high, resulting in a substantial increase in GPU revenues for the quarter even as the units declined.

I don't know about you but Mercury numbers aren't aligning with what she stated. And we know when they stated average desktop selling prices, they weren't talking about AMD.

The whole thing is a facade.

If they want to run numbers like this was accounting, they really should be somewhere else.

 

[firstminion]

There internal shipments right?

Mercury numbers should show that?

Do you see that here?

Highest desktop channel since 2014 , look at desktop discrete, 22.8% that is lower than 2015 q2!
Now with the drop in overall shipments you think, AMD's shipments have increased?



I don't know about you but Mercury numbers aren't aligning with what she stated. And we know when they stated average desktop selling prices, they weren't talking about AMD.

The whole thing is a facade.

If they want to run numbers like this was accounting, they really should be somewhere else.

Mercury could be less precise than AMD's own accounting and absolute numbers of 2016Q2 are probably higher than 2015Q2.

Does it really matter anyway? Are we really complaining that "AMD sold more, but not that much more"?