East said ARM's objectives with regard to AMD are broad and would not only focus on Mali graphics processors. Indeed AMD has its own graphics operation having bought ATI a few years ago.
"At the time ATI was actually an ARM licensee for some of its work in mobile applications so AMD did technically become an ARM licensee." Qualcomm then bought the mobile graphics division from AMD for $65 million.
AMD is now a fabless chip company having spun off its manufacturing operations to form the beginnings of Globalfoundries Inc., a foundry with which ARM also works closely.
As a fabless company working with OEMs AMD is obliged to think about the lowest cost and lowest power consumption in terms of the hardware-software combinations it provides and enables.
If negotiations were starting today they would probably focus on ARM's forthcoming Cortex-A15 multicore-capable processor core. But East declined to rule out the possibility of licensing Cortex-A8 or Cortex-A9 to AMD.
I am not an expert, but as end user I know that Intel is very fast, AMD is fast, my ARM tablets/phones are slow.
To race the Indy 500 I need a 12 cylinders engine and a lot of fuel, to commute to/from work the old 4 cylinders does the job using 1 Gal for 25 miles.
What about a new architecture? Also the Arm concept is old! Can it be enhanced to a high performance device and how many programs the will be available for it? is this a PowerPC repeat? is there a unified design standard? and so on.
I believe that he chip technology is the paramount factor in power reduction.
I think AMD, that is geared to offer high performance devices, will not be profitable making he ARM unless they capture a vary good chunk of the market.
Smart move by ARM: trying to put a foot into high end computing through a proxy :-) As for AMD, it does not make sense for them to drop x86, at least not for now - why should they?
By the way, I do not buy into the immortality of x86. I could well see a world without it, and during my life time. Efficient compilers, emulators and thin-client computing could smooth the transition.
ChromeOS might be viable as on a Net machine where one does all of the normal financial transactions and is fairly virus-proof. Beyond that, it's got to be the next version of Android. I mean, how much would it take to make Android a general desktop OS as slick as Win7 or OSX?
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.