[Digitimes] AMD expected to reach 40% GPU share in 6 months

[HurleyBird]

Not surprising given the strength of next gen consoles, and artificial demand inflation due to cryptocoin mining.

 

[wand3r3r]

They have some room to go still:

http://www.cgw.com/Press-Center/News/2014/Add-in-Board-Market-Up-in-Q3-Nvidia-Gains-Market.aspx

Is this the Q3 from last summer July-Sept 2013?

 

[OCGuy]

Not surprising given the strength of next gen consoles, and artificial demand inflation due to cryptocoin mining.

I'm not following...

 

[ShintaiDK]

Not surprising given the strength of next gen consoles, and artificial demand inflation due to cryptocoin mining.

Crypto is essentially US only. A relatively very small market.

If you want to count APUs, then AMD is down to around 20% and no chance at all to each 40%.

 

[Pandamonia]

LOL, AMD's discrete graphics market share is already about 38% (compared to 62% for NVIDIA), so that's not saying much. And with NVIDIA's recent aggressive price cuts, I don't think that AMD will be able to gain any significant share. Wait six months and we will find out.

Aggressive price cuts? what planet are you on?

£400 for a 780. £540 for a 780 Ti. £850 for Titan.

290 is £310, 290X is £400

Nothing aggressive about any of those prices except AMD

 

[ICDP]

Is this the Q3 from last summer July-Sept 2013?

Indeed it is, so the market share OCGuy posted is from before AMD made this statement. Nice bit of sleuthing there.

AMD's statement was made in mid Q4 and should by by that context mean Q4 2013 and Q1 2014. Not saying they will reach their goal but simply pointing out OCGuy is very premature with his post.

 

[OCGuy]

Indeed it is, so the market share OCGuy posted is from before AMD made this statement. Nice bit of sleuthing there.

AMD's statement was made in mid Q4 and should by by that context mean Q4 2013 and Q1 2014. Not saying they will reach their goal but simply pointing out OCGuy is very premature with his post.

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

 

[bystander36]

Aggressive price cuts? what planet are you on?

£400 for a 780. £540 for a 780 Ti. £850 for Titan.

290 is £310, 290X is £400

Nothing aggressive about any of those prices except AMD

In the US, AMD prices on the 290 are about the same as the 780.

 

[96Firebird]

Aggressive price cuts? what planet are you on?

£400 for a 780. £540 for a 780 Ti. £850 for Titan.

290 is £310, 290X is £400

Nothing aggressive about any of those prices except AMD

His post was from early November, when Nvidia had recently cut prices to the 780 and 770.

 

[ICDP]

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

So instead of saying, sorry for resurrecting an old thread with FUD, you simply present even more FUD. Just admit you were wrong and move on, no shame in it.

Come back in 3-4 months with the latest JPR data is available and we can discus if AMD are full of crap or not.

 

[OCGuy]

So instead of saying, sorry for resurrecting an old thread with FUD, you simply present even more FUD. Just admit you were wrong and move on, no shame in it.

Come back in 3-4 months with the latest JPR data is available and we can discus if AMD are full of crap or not.

I posted data that was just released, as well as another source that is current. Ad hominem attacks and emotional language are not necessary to prove a point.

 

[ICDP]

I posted data that was just released, as well as another source that is current. Ad hominem attacks and emotional language are not necessary to prove a point.

Lighten up, it's called banter and should not be taken seriously. You did notice the laughing smiley implying "this is a joke, do not take this seriously"?

 

[Skurge]

I posted data that was just released, as well as another source that is current. Ad hominem attacks and emotional language are not necessary to prove a point.

Steam survey is not reliable data. I think I will wait for when JPR data is available again.

 

[BallaTheFeared]

Steam survey is not reliable data. I think I will wait for when JPR data is available again.

That should be more favorable for AMD since a lot of their cards are not being purchased for gaming and thus probably will never show up on steam surveys.

 

[Gikaseixas]

Crypto is essentially US only. A relatively very small market.

In the US, AMD prices on the 290 are about the same as the 780.

People will try really hard to downplay anything positive about AMD.
What's going on here?
Is the US a big enough market or is it insignificant in a global perspective??? Make up your minds LOL

 

[Gikaseixas]

That should be more favorable for AMD since a lot of their cards are not being purchased for gaming and thus probably will never show up on steam surveys.

Regardless...
... market share does not care if a card was bought for mining or gaming purposes or does it?
Right so JPR data is more reliable, i think we all agree

 

[BallaTheFeared]

Regardless...
... market share does not care if a card was bought for mining or gaming purposes or does it?

No but it's disingenuous to what the market share is supposed to represent (gaming), otherwise they'd include professional solutions in it.

I couldn't care less either way, just pointing out what I believe to be true.

 

[OCGuy]

That should be more favorable for AMD since a lot of their cards are not being purchased for gaming and thus probably will never show up on steam surveys.

AMD already had X amount of R9 chips to send out. They couldn't move more than they had. They put more 28nm orders in @ TSMC for Q1, but:

1. The top cards make up less than 1%, so I don't see that moving the peg as much as claimed in OP

2. Nobody knows how many total Hawaii cards there were to actually be sold in Q4, even including the mining craze. Lack of inventory @ etailers only gives demand, not supply.

Either way Steam is a fairly good sample size, but more data will come.

Q4 was great for AMD as a company, they sold every high-margin chip they could. But that is not what is being discussed in this thread.

 

[bystander36]

People will try really hard to downplay anything positive about AMD.
What's going on here?
Is the US a big enough market or is it insignificant in a global perspective??? Make up your minds LOL

My comment was in response to the very rude one made in my quote about how things are different on his planet.

I'm on his planet, and in the US, it is different than he claimed it was for him.

So, the point is, depending on where you are, the price differences may be small.

 

[blastingcap]

I'm not disagreeing with your software/driver quality argument; this been one of the greatest strengths of Nvidia since the first Detonator drivers, but sometimes I wonder if AMD's past reputation is perpetuated by hearsay rather than actual facts. I've been using ATI/AMD products for years, and to be fair, I've never really had any problems with them. Now, that doesn't mean those bugs don't exist, but when I ask individuals about their poor experiences with AMD software, very few can actually give me any specific answers. Just as an example, my brother swears up and down by Nvidia, which is fine, but when he starts commenting on why he doesn't use AMD products and how poor AMD's drivers are in comparison to Nvidia, I feel like all I get is a made-up story. Whenever I've ever talked to anyone in person, all I ever seem to hear is someone's friend's experiences, or about something they read online; never anyone's own, personal experience.

So I have to wonder just based on my own observation, are some of these issues people are having with AMD simply just, well, fabricated?

The short answer:

AMD driver quality:

Single display up to 2560x1600 with one or more GPUs - good
Single display up to 2560x1600 with multiple GPUs - good
4K or multimonitor/Eyefinity with single GPU - good
4K or multimonitor/Eyefinity with multiple GPUs - good starting with R9 290 series for DX11 games (not sure about pre-DX11 games), bad for everything prior to that though the company has promised to fix this

 

[blastingcap]

AMD already had X amount of R9 chips to send out. They couldn't move more than they had. They put more 28nm orders in @ TSMC for Q1, but:

1. The top cards make up less than 1%, so I don't see that moving the peg as much as claimed in OP

2. Nobody knows how many total Hawaii cards there were to actually be sold in Q4, even including the mining craze. Lack of inventory @ etailers only gives demand, not supply.

Either way Steam is a fairly good sample size, but more data will come.

Q4 was great for AMD as a company, they sold every high-margin chip they could. But that is not what is being discussed in this thread.

"The top cards" meaning what, R9 290/X? The R9 280X and even lower cards have also received more demand so sales are up for more than just the top 1%.

 

[Gikaseixas]

The short answer:

AMD driver quality:

Single display up to 2560x1600 with one or more GPUs - good
Single display up to 2560x1600 with multiple GPUs - good
4K or multimonitor/Eyefinity with single GPU - good
4K or multimonitor/Eyefinity with multiple GPUs - good starting with R9 290 series for DX11 games (not sure about pre-DX11 games), bad for everything prior to that though the company has promised to fix this

I had many problems with ATI in the past and used Nvidia for years without a single driver related issue. When AMD launched the HD 5870 i bought it and fell in love with the card, since then i'm having trouble free experience at a cheaper prices. I'd say that people really need to actually buy, use and see for themselves before spreading false info around the net.
It's old, update your opinions (not you blastingcap)

 

[f1sherman]

The quarter in general

JPR found that AIB shipments during Q3, 2013 behaved according to past years with regard to seasonality, but the increase was less than the 10-year average.
AIB shipments increased 3.9% from the last quarter (the 10-year average is 12%).

• Total AIB shipments increased this quarter to 14.5 million units.

• AMD’s quarter-to-quarter total desktop AIB unit shipments decreased 3%.

• Nvidia’s quarter-to-quarter unit shipments increased 8%.

• Nvidia continues to hold a dominant market share position at 64.5%.

• Figures for the other suppliers were flat to declining.


http://jonpeddie.com/publications/add-in-board-report/

 

[ICDP]

The quarter in general

JPR found that AIB shipments during Q3, 2013 behaved according to past years with regard to seasonality, but the increase was less than the 10-year average.
AIB shipments increased 3.9% from the last quarter (the 10-year average is 12%).

• Total AIB shipments increased this quarter to 14.5 million units.

• AMD’s quarter-to-quarter total desktop AIB unit shipments decreased 3%.

• Nvidia’s quarter-to-quarter unit shipments increased 8%.

• Nvidia continues to hold a dominant market share position at 64.5%.

• Figures for the other suppliers were flat to declining.


http://jonpeddie.com/publications/add-in-board-report/

That data is from Q3 2013 and not in any way relevant to the point of the OP.

The OP was made on the 4th November 2013 stating AMD expected to reach 40% GPU share in 6 months. So the claim is relating to an at the date of the OP future point in time from Nov 2013 up to May 2014. So I ask again, why are last years JPR figures from July - Sept 2013 relevant and being posted to disprove this prediction?

Am I missing something here? Either I am being very stupid, or the information you presented is the same past and therefore irrelevant figures that OCGuy posted.

Example: If I say, "I think it will be wet this coming spring". Would you respond by showing me the previous summer's weather reports and saying, "these weather reports from last summer prove you are wrong"? I don't think you would because it bears absolutely zero relevance to my prediction.

AMD may not reach this target, but we cannot start claiming one way or the other until the middle of 2014.

 

[f1sherman]

It's just the JPR AIB report.
It gets posted here periodically, usually in it's own thread.

Click away if it's not in any way relevant