I keep coming back to this statement on an AMD CES slide: "(Gaming) represents a massive market opportunity for AMD".
From a TR Report article on the CES presentation: "Su described graphics as one of AMD's "crown jewels" that ends up being the "centerpiece of our roadmap."
That's straight from the horses mouth and not even slightly ambiguous, it tells us exactly where gaming figures in AMD's priorities ... the gaming market is flat out AMD's no. 1 priority, bar none. In the context of this OP the importance of this reality cannot be overstressed. Please keep it in mind as you read this post.
Sewing up both upcoming next gen consoles and Steam Box would obviously be a top priority in addressing this 'massive market opportunity'.
The most reliable leaks to date point to an AMD CPU and GPU in the Xbox 720 and PS4. Kaveri is CES road mapped to be retail available as the Xbox 720 and PS4 release. Hence those architectural elements were available for consideration when Microsoft and Sony were rumored to have reconsidered their console plans and ended up going with an AMD solution. Console designers want as much future proofing as possible and Kaveri technologies, extensive HSA implementation, unified address space, common memory pool, GCN 2.0 and further refined 'Steamroler' cores are just that. Why wouldn't Microsoft and Sony go for at least those advanced technologies if they were available and there was reasonable certainty it could meet 2013 console release mass production targets?
AMD would necessarily be heavily involved with the entire next gen game development stack for Sony and Microsoft.
Why wouldn't AMD have modified Kaveri, on the same design/fabrication timeline as the console chips, to optimize it's ability to play next gen games? There's a very substantial synergy there to be taken advantage of that also coincides with AMD's top priority.
There are a number of rumors Steamroller (FX) has been pushed back, or even terminated, along with the whole standalone CPU line. Why not? There are no 'massive' market opportunities going head to head with Intel in those market segments. Micro servers are AMD's focus now, that means ARM and Jaguar based chips. RIP FX and Opteron?
Considering what came out at CES, AMD dropping their standalone CPUs to concentrate on rapid iterations of their gaming optimized APUs gaming makes far more sense.
It is likely Kaveri will be 'refreshed' sooner than later and that refresh will be a fully optimized stacked chip gaming monster. It is also likely the 9xxx graphics boards will be the last of their kind and interface optimized for AMD APUs.
It's like a 'perfect storm' of synergies AMD can take advantage of, and they manifestly are prioritized for and putting huge resources behind doing just that. And saying so publicly.
Two years from now it is well within the realm of possibilities AMD will have a line of gaming APUs and AIBs neither Intel or Nvidia can compete with.
Which brings me to the thread title.
Steam Box is the wild card. Considering Gabe's record to date, when it becomes available Valve's own Steam Box and the Steam Box branded ecosystem can be expected to explode and within a few years have numbers in the tens of millions, a major gaming market.
If Valve goes with an Intel/Nvidia solution for their Steam Box console, Nvidia in particular retains a competitive toehold in computer gaming it can expand on.
But if Valve goes with an AMD solution, particularly one already optimized for next gen games to get the developers on board to provide Linux ports, then what future does Nvidia have in gaming graphics?
But Gabe is cagey. So far the only name he's dropped is Nvidia's. That he looks at Steam Box from a long term strategic point of view is a given. How does that play out to decide which CPU/GPU solution he ultimately goes with? No pay to predict. He may see AMD cornering the console and PC gaming markets TOO thoroughly. He may find AMD synergies just too tempting to resist. What is reasonable certain is his choice will have substantial consequences.
If he goes AMD, Nvidia would have no seat at the next gen consoles or Steam Box Linux game development table at all.
That leaves x86 PC games and AMD would by default already be solidly present there from the consoles and even more so if developers had a focus on optimizing for AMD's upcoming Gaming Evolved APU lines. Doesn't leave much space for Nvidia even at that last development table.
IF this is how it all plays out game developers and publishers will be able to read those signs easily enough and see the gaming future belongs to AMD and allocate their resources and manpower accordingly.
That is my reasoning for positing Valve might ultimately determine Nvidia's fate in the gaming GPU space.
From a TR Report article on the CES presentation: "Su described graphics as one of AMD's "crown jewels" that ends up being the "centerpiece of our roadmap."
That's straight from the horses mouth and not even slightly ambiguous, it tells us exactly where gaming figures in AMD's priorities ... the gaming market is flat out AMD's no. 1 priority, bar none. In the context of this OP the importance of this reality cannot be overstressed. Please keep it in mind as you read this post.
Sewing up both upcoming next gen consoles and Steam Box would obviously be a top priority in addressing this 'massive market opportunity'.
The most reliable leaks to date point to an AMD CPU and GPU in the Xbox 720 and PS4. Kaveri is CES road mapped to be retail available as the Xbox 720 and PS4 release. Hence those architectural elements were available for consideration when Microsoft and Sony were rumored to have reconsidered their console plans and ended up going with an AMD solution. Console designers want as much future proofing as possible and Kaveri technologies, extensive HSA implementation, unified address space, common memory pool, GCN 2.0 and further refined 'Steamroler' cores are just that. Why wouldn't Microsoft and Sony go for at least those advanced technologies if they were available and there was reasonable certainty it could meet 2013 console release mass production targets?
AMD would necessarily be heavily involved with the entire next gen game development stack for Sony and Microsoft.
Why wouldn't AMD have modified Kaveri, on the same design/fabrication timeline as the console chips, to optimize it's ability to play next gen games? There's a very substantial synergy there to be taken advantage of that also coincides with AMD's top priority.
There are a number of rumors Steamroller (FX) has been pushed back, or even terminated, along with the whole standalone CPU line. Why not? There are no 'massive' market opportunities going head to head with Intel in those market segments. Micro servers are AMD's focus now, that means ARM and Jaguar based chips. RIP FX and Opteron?
Considering what came out at CES, AMD dropping their standalone CPUs to concentrate on rapid iterations of their gaming optimized APUs gaming makes far more sense.
It is likely Kaveri will be 'refreshed' sooner than later and that refresh will be a fully optimized stacked chip gaming monster. It is also likely the 9xxx graphics boards will be the last of their kind and interface optimized for AMD APUs.
It's like a 'perfect storm' of synergies AMD can take advantage of, and they manifestly are prioritized for and putting huge resources behind doing just that. And saying so publicly.
Two years from now it is well within the realm of possibilities AMD will have a line of gaming APUs and AIBs neither Intel or Nvidia can compete with.
Which brings me to the thread title.
Steam Box is the wild card. Considering Gabe's record to date, when it becomes available Valve's own Steam Box and the Steam Box branded ecosystem can be expected to explode and within a few years have numbers in the tens of millions, a major gaming market.
If Valve goes with an Intel/Nvidia solution for their Steam Box console, Nvidia in particular retains a competitive toehold in computer gaming it can expand on.
But if Valve goes with an AMD solution, particularly one already optimized for next gen games to get the developers on board to provide Linux ports, then what future does Nvidia have in gaming graphics?
But Gabe is cagey. So far the only name he's dropped is Nvidia's. That he looks at Steam Box from a long term strategic point of view is a given. How does that play out to decide which CPU/GPU solution he ultimately goes with? No pay to predict. He may see AMD cornering the console and PC gaming markets TOO thoroughly. He may find AMD synergies just too tempting to resist. What is reasonable certain is his choice will have substantial consequences.
If he goes AMD, Nvidia would have no seat at the next gen consoles or Steam Box Linux game development table at all.
That leaves x86 PC games and AMD would by default already be solidly present there from the consoles and even more so if developers had a focus on optimizing for AMD's upcoming Gaming Evolved APU lines. Doesn't leave much space for Nvidia even at that last development table.
IF this is how it all plays out game developers and publishers will be able to read those signs easily enough and see the gaming future belongs to AMD and allocate their resources and manpower accordingly.
That is my reasoning for positing Valve might ultimately determine Nvidia's fate in the gaming GPU space.
Last edited:
Facebook
Twitter