Per the CES AMD roadmap Kaveri will be available Q4, 2013. Ergo, the design was finalized when Microsoft and Sony are rumored to have 86ed their existing next gen plans for something else. Apparently AMD literally made Microsoft and Sony an offer they couldn't refuse.
Per AMD"s CES slideshow - "(Gaming) represents a massive market opportunity for AMD".
That's a VERY strong statement from AMD. But what other large and lucrative market segment offers such opportunity? It's the only large market AMD has the slightest chance to all but totally own. It's exactly what the slide says - a 'massive opportunity'. And every indication is they are intensely focused on, and putting the necessary resources into, taking full advantage of that opportunity.
Kaveri - Steamroller cores, extensive HSA implementation, unified address space, fully shared memory, GCN 2.0 ... add in AMD's 'mix and match' interface ... what else comes within a country mile of meeting next gen console performance/power/future proofing needs? Including Steam Box's needs?
Based on a working assumption AMD is getting their Kaveri architecture into the Xbox 3, PS 4 and Steam Box, there are wide ranging implications.
A deep and comprehensive development ecosystem likely to grow up around that architecture. All console and PC game engines likely to key on that architecture. The entire Steam Box Linux based console and PC ecosystem would grow around that architecture if it was chosen because AMD is almost certain to craft custom Linux drivers only for that particular APU family. It seems likely to be a de-facto pre-emption of Intel and Nvidia from future Linux development.
In a year, with all next gen and Steam Box Linux games tuned to it's architecture, Kaveri is likely to become the hot gaming chip and similarly architecture 8xxx boards to be the go to choice for gamers looking for more power. Under this scenario Intel's single core performance edge would not be coded for and not be taken advantage of nearly to the extent that Kaveri's Steamroller cores, HSA features, common memory pool, unified address space and GCN 2.0 architecture would be. Kaveri would just provide a far better gaming experience than anything on offer from Intel. It wouldn't take long for this to spread in the gaming community. A year from now is likely to see Kaveri being in very high demand. Two years from now Intel is likely to be all but irrelevant to computer gamers. It's a matter of economics. How expensive an Intel/Nvidia rig would it take to match the performance of a $130 Kaveri chip/$100 HD8750 board combo on next gen games? It's probably safe to say at least twice the $$$ for the same performance.
AIB graphics boards based on Kaveri GCN 2.0 would have an innate advantage over Nvidia's architecture as all gaes would be coded for that architecture and AMD, now intensely focused on gaming, would have the inside track on providing tip notch drivers day 1.How long would it take before Nvidia is relegated to second class status? How could they be price/performance competitive with AMD when AMD is holding a handful of aces? How can Nvidia compete in a professional market rapidly adopting the HSA architecture and having no OEM or AIB market sales of significance to underwrite the costs of developing Professional Graphics boards?
The professional graphics players, seeing the writing on the wall, are likely to start porting their engines to AMD's HSA architecture sooner than later. It's probably in progress now. How would Nvidia not becomes a minor player? I give JHH his due. He's a hell of a fighter. But forces are in motion beyond his control. I just don't see how Nvidia survives in the x86 space when they are shut out of the architecture.
Sure appears the potential is there for Kaveri architecture to largely determine the direction of the video card markets going forward, in essence, coming to 'rule' the consumer and professional video card markets.
Per AMD"s CES slideshow - "(Gaming) represents a massive market opportunity for AMD".
That's a VERY strong statement from AMD. But what other large and lucrative market segment offers such opportunity? It's the only large market AMD has the slightest chance to all but totally own. It's exactly what the slide says - a 'massive opportunity'. And every indication is they are intensely focused on, and putting the necessary resources into, taking full advantage of that opportunity.
Kaveri - Steamroller cores, extensive HSA implementation, unified address space, fully shared memory, GCN 2.0 ... add in AMD's 'mix and match' interface ... what else comes within a country mile of meeting next gen console performance/power/future proofing needs? Including Steam Box's needs?
Based on a working assumption AMD is getting their Kaveri architecture into the Xbox 3, PS 4 and Steam Box, there are wide ranging implications.
A deep and comprehensive development ecosystem likely to grow up around that architecture. All console and PC game engines likely to key on that architecture. The entire Steam Box Linux based console and PC ecosystem would grow around that architecture if it was chosen because AMD is almost certain to craft custom Linux drivers only for that particular APU family. It seems likely to be a de-facto pre-emption of Intel and Nvidia from future Linux development.
In a year, with all next gen and Steam Box Linux games tuned to it's architecture, Kaveri is likely to become the hot gaming chip and similarly architecture 8xxx boards to be the go to choice for gamers looking for more power. Under this scenario Intel's single core performance edge would not be coded for and not be taken advantage of nearly to the extent that Kaveri's Steamroller cores, HSA features, common memory pool, unified address space and GCN 2.0 architecture would be. Kaveri would just provide a far better gaming experience than anything on offer from Intel. It wouldn't take long for this to spread in the gaming community. A year from now is likely to see Kaveri being in very high demand. Two years from now Intel is likely to be all but irrelevant to computer gamers. It's a matter of economics. How expensive an Intel/Nvidia rig would it take to match the performance of a $130 Kaveri chip/$100 HD8750 board combo on next gen games? It's probably safe to say at least twice the $$$ for the same performance.
AIB graphics boards based on Kaveri GCN 2.0 would have an innate advantage over Nvidia's architecture as all gaes would be coded for that architecture and AMD, now intensely focused on gaming, would have the inside track on providing tip notch drivers day 1.How long would it take before Nvidia is relegated to second class status? How could they be price/performance competitive with AMD when AMD is holding a handful of aces? How can Nvidia compete in a professional market rapidly adopting the HSA architecture and having no OEM or AIB market sales of significance to underwrite the costs of developing Professional Graphics boards?
The professional graphics players, seeing the writing on the wall, are likely to start porting their engines to AMD's HSA architecture sooner than later. It's probably in progress now. How would Nvidia not becomes a minor player? I give JHH his due. He's a hell of a fighter. But forces are in motion beyond his control. I just don't see how Nvidia survives in the x86 space when they are shut out of the architecture.
Sure appears the potential is there for Kaveri architecture to largely determine the direction of the video card markets going forward, in essence, coming to 'rule' the consumer and professional video card markets.
Last edited: