AMD could be profitable in 2 years.

[Mondozei]

The article is here.

Since paywalls are terrible and a stain on humanity, we also have a reddit transcript.

This on the heels that Jon Peddie claims that AMD has gained some dGPU market share(just slightly) in Q3. With the perception of AMD being better for DX12 and VR slowly sinking in - and the dubious nonsense that AMD drivers are somehow "bad" slowly withering - it remains to be seen if this momentum can be kept up.

Normally, I take these kinds of prognostications skeptically but in this instance AMD does look fairly good. 2016/2017 shapes up to be the best period they've had in terms of roadmap attractiveness in many years.

 

[Skurge]

They seem to have stronger more focused leadership in Lisa Sue and Raja Kaduri for RTC.

We will see their changes reflect in 2016/2017.

 

[jpiniero]

Yeah no. Since you posted it in the GPU forum, I'll ignore the CPU business, but it's not like the GPU business is all that healthy. The GPU business has never been that profitable, but that's certainly much better than the CPU has done. It's still a declining business and for it to not be a bigger problem they would need to gain share with non-closeout products so they make anything about it.

My 'fear' is that Arctic Islands is HBM2 only, and if HBM2 were to be delayed AMD could be faced with having to compete with Pascal w/ GDDR5X with nothing really until 2017. Even if that Pascal really isn't that much faster it's going to be really hard for sales to not start to plummet again. I don't think they want to get in a position where they have to sell Furies for $399, they would probably lose money.

This is where the R&D cuts would start to show up. It's not right away.