AMD VEGA confirmed for 2017 H1

[N19h7m4r3]

I've seen you post this loads of times in this thread, is this a running joke that I'm missing or do you have a source saying the Vega design isn't for gaming?

He's parroting what Doom has been posting for nearly a year.

Also "Vega is for VR", yeah cause there are no VR games.

 

[Confused Stu]

He's parroting what Doom has been posting for nearly a year.

Also "Vega is a VR", yeah cause there are no VR games.

Thanks mate. Was pretty sure it was a joke that I missed (there are a lot of posts in this thread to keep track of!) but just started to wonder after seeing it so often.

Cheers for clarifying.

 

[D.P.]

AMD either need to outperform the competition, or outprice. (is this a word?)

If they plan on releasing similar performance, for similar prices, I've no idea how they expect to gain a large percentage of market share.

They may not have that much flexibility in pricing though once you factor in HBm2 costs, and if the rumors of short supply are true there is no way for AMD to actually increase market share substantially regardless of price.

I think the bgger picture is how does AMD intend to become a more profitable and healthy company that can make significant R&D investments, diverse product offerings, happy sharehodlers, pay off their debts. Market share is one aspect. If AMD can slowly push their market share up to 30-35% of discrete that would still be great with console sales, APUs, CPUs, HPC, deep learning, automotive/AV. A high profit-margin 30% market share of graphics cards with millions of low profit console sales would work out very well. 50% market share sold with bare minimum margins would destroy AMD.

 

[LoadsaMoney]

I've seen you post this loads of times in this thread, is this a running joke that I'm missing or do you have a source saying the Vega design isn't for gaming?

http://wccftech.com/amd-vega-mi-25-instinct-gpu-accelerator/



Polaris are their cards for VR.

AMD Polaris GPUs aiming to offer minimum spec VR under the $349 mark
Before we go any further, a little bit of context is in order. Running virtual reality headsets such as the Oculus Rift and HTC Vive requires quite a bit of horsepower, which in the case of the Oculus Rift is 2k resolution at 90 frames per second. This graphical horsepower can be produced by GPUs equivalent (or more) in power to the GTX 970 or the Radeon R9 290. Both these GPUs cost $349. So if we take a look at the sales data of these GPUs as well as the GPUs above this mark, we will be able to calculate something called the total available market for VR – which is the number of users that can actually run any VR product on these headsets. AMD believes, that Polaris architecture will be efficient enough to address this problem:

Now I would like to address a different challenge for us – and that is what is called a Total Available Market. One of the issues we have is the minimum spec for the PCs which will run the Occulus and the HTC headset ant 90 fps and 2k resolution. Now to do that you need either a Radeon 290x or the GTX 970 both of which retail for $349. The challenge that we have is if you look at the total numbers of these GPUs that have been sold, according to JPR, that’s an install base of just 7.5 Million units. Now that’s an issue because it means you can only sell 7.5 Million of anything – because that’s the number that can run those headsets. I am very pleased to tell you that we have invented something called Polaris which we think will address this problem. – AMD’s Roy Taylor

http://wccftech.com/amd-polaris-architecture-vr-minimum-spec/

 

[BriT]

Thought wcc was a load of rubbish

 

[Mtom]

They may not have that much flexibility in pricing though once you factor in HBm2 costs, and if the rumors of short supply are true there is no way for AMD to actually increase market share substantially regardless of price.

I think the bgger picture is how does AMD intend to become a more profitable and healthy company that can make significant R&D investments, diverse product offerings, happy sharehodlers, pay off their debts. Market share is one aspect. If AMD can slowly push their market share up to 30-35% of discrete that would still be great with console sales, APUs, CPUs, HPC, deep learning, automotive/AV. A high profit-margin 30% market share of graphics cards with millions of low profit console sales would work out very well. 50% market share sold with bare minimum margins would destroy AMD.

Well we've just seen a pic comparing Fiji with HBM and Hawaii with GDDR5, and the price difference was small.
As multiple articles wrote the 16-20000 piece at start is normal at high end GPUs so it's not really a problem.

Also you might missed the news, but AMD is at ~ 30% GPU market share for a year now. They claimed a good chunk before Polaris launch.

 

[labbby]

Thought wcc was a load of rubbish

It is unless someone is trying to use it to back up a point they have.. It's funny some people on here link wcc to back up their arguments but also tell us that it's a load of rubbish.. but that's when wcc is saying the opposite to their argument

 

[LoadsaMoney]

Thought wcc was a load of rubbish

What about AMD themselves, VEGA is at the bottom, MI25


http://instinct.radeon.com/en-us/about/

They are slapping an air/liquid cooler onto it, upping its clocks, and selling it to those who want to use it for gaming.

 

[D.P.]

Well we've just seen a pic comparing Fiji with HBM and Hawaii with GDDR5, and the price difference was small.
As multiple articles wrote the 16-20000 piece at start is normal at high end GPUs so it's not really a problem.

Also you might missed the news, but AMD is at ~ 30% GPU market share for a year now. They claimed a good chunk before Polaris launch.

No, the price difference was huge, $42 which will massively cut margins. NVidia and AMD might get 40-80 margin per GPU out of which they would have to pay R&D and other none-BoM costs. Removing $40 from that is not sustainable.
People always wonder why some GPUs use a cheap VRm/capacitor/screw/fan. Well, AMD and NVidia need to be saving the 20cents here, the 5cent there, 30 cents there t in order to preserve margins. A $42 difference is massive, so the card will have to be sold at least 40 more. And that is ignoring things like yield issues, how many successful HBM+interposer+GPU DIes can AMD deliver vs simply sending a GPU die etc. We know form Fiji that AMD suffered high failure rates due to the complexity with the interposer.

EDIT: yes, I know AMD are at about 30% market share, that is just my point they now need to concentrate on profits.

 

[MichaelM]

So they showed Vega running Battlefront and Doom to prove that it is not for gaming?

 

[KillBoY_UK]

As multiple articles wrote the 16-20000 piece at start is normal at high end GPUs so it's not really a problem.

We had similar stories about the Fury X, and looks what happen VERY low stock levels on release the world over. I think OCUK had less than 10 on the day and they are prob the people whoweree able to get the most stock in the UK.

 

[LoadsaMoney]

So they showed Vega running Battlefront and Doom to prove that it is not for gaming?

It can be used for gaming as well.

 

[N19h7m4r3]

It's because unlike NVIDIA, AMD make one unified GPU design. The Vega 10 GPU is used in Instinct ( Tesla ), Radeon Pro ( Quadro ), and Radeon ( GeForce ).

So it's a big chip, that'll do it all really; just like AMD has been doing since the 7970 launch.
Nothing has changed since then for AMD, bar DP performance slowly being cut.

It can be used for gaming as well.

 

[AlamoX]

It's because unlike NVIDIA, AMD make one unified GPU design. The Vega 10 GPU is used in Instinct ( Tesla ), Radeon Pro ( Quadro ), and Radeon ( GeForce ).

So it's a big chip, that'll do it all really; just like AMD has been doing since the 7970 launch.
Nothing has changed since then for AMD, bar DP performance slowly being cut.

and yet that same chip have much more performance than Gforce equivalent, when not bottlenecked by 3rd party Software/API.

 

[Jono8]

It would be interesting to know just how far along Nvidia are with Volta. As now this is looking like it won't be out till at least June, Nvidia could make the shelf life of this (at least at high prices) very short if they release Volta before the end of the year! This is assuming Vega is in the 1080-1080ti performance range.

 

[D.P.]

It would be interesting to know just how far along Nvidia are with Volta. As now this is looking like it won't be out till at least June, Nvidia could make the shelf life of this (at least at high prices) very short if they release Volta before the end of the year! This is assuming Vega is in the 1080-1080ti performance range.

2017Q4 has a chance with a 2080/2070, Volta titan in Q1 and 2080ti in Q1-Q3 2018

 

[N19h7m4r3]

2017Q4 has a chance with a 2080/2070, Volta titan in Q1 and 2080ti in Q1-Q3 2018

They could also wait until GDDR6 is out and available since GDDR5X is nearly at its limit.

 

[D.P.]

They could also wait until GDDR6 is out and available since GDDR5X is nearly at its limit.

GDDR6 For the the Titna and 2080ti looks likely, no reason not to use GDDR5x for the midrange really.

 

[N19h7m4r3]

GDDR6 For the the Titna and 2080ti looks likely, no reason not to use GDDR5x for the midrange really.

Could be, they can try and squeeze everything they can out of it before replacing it all; then do a "Refresh" with GDDR6 later on when it's more readily available.

 

[AlamoX]

i still don't understand the ppl who think that Vega will be like 600-700$ GPU, it simply won't, because more important than performance is market segment, AMD cannot go from 229$ product to 699$ product, and lose on volume segments...
AMD Will have a 300-350$ Vega and a 450-500$ Vega, this is not prediction, this is a must for AMD, and they are not stupid enough to bypass that.
once that is done, they can find some extra value to add to bring the premium margins up, so the HBM2 and interposer cost have been factured in long before this, and the Vega Sku that sits in each segment it would have the regular 30ish% margins AMD often opte for.