'Apple would be crazy not to use IBM's new PPC chip.' This is what Peter Glaskowsky, director of the prestigious and authoritative The Microprocessor Report, thinks about the possibility that the 64-bit Power4 that IBM is about to present could end up inside the Mac.
Glaskowsky, as well as director of what has long been one of the main media in the sector, also the organizer of the Microprocessor Forum, an event that is held every year in San Jos and that presents a glimpse into the future of the semiconductor and CPU field for PC.
Right at the Microprocessor Forum IBM should for the first time expose the main features of the 64-bit Power4, a derivative of the original Power4 whose target are the high-end servers and which will bring a series of innovations on the PowerPc platform that will translate into an increase very sensitive in computing skills.
'Apple' told EETimes Glaskowsky 'could for the first time have a 64-bit processor available which, moreover, will have significant power, even if at this moment I am not able to say which, which will surely be vertices of all CPUs. Apple would be foolish not to take advantage of this option '
Of course, Apple will have a lot of work to do to integrate Power4 into their production line. Despite the fact that it can work transparently in 32 bit, 64 bit compatibility is the most interesting aspect in perspective. And to bring the operating system to 64 bit it will take some time; even more is needed for applications to be brought to 64 bit. "The effort to be able to fully use the processor will be great," says Glaskowsky.
Recall that according to some information collected in recent weeks the Power4 includes a vector unit that probably is none other than AltiVec and has a superscalar eight-stage data assembly line and supports symmetric multiprocessing.
According to many sources, if Apple really adopts the Power4, the first Macs with this type of processor are likely not to arrive before the end of 2003.
On the other hand, there are the perplexities of the developers who should optimize the code of some "important" applications in order to make the most of the CPU characteristics. It would be yet another evolution after the not painless transition (in terms of time and costs ) to Mac OS X.