A rumor it is; but the concept is something to think about. If you take a couple of things into consideration, such as Apples commitment to multi threading and the use of vector operations, then you could see where this is a very real possibility. Now this could be a logn term joke who knows but it is none the less completely possible, though as mentioned before I really thing such a machine should be targeted at the low pwoer market and not the PowerBook market.
To keep things truely Symetric they would have to have a vector unit and FPU attached to each processor. So you would never be doing vector operation any greater than was implemented for each processor. You can not think of this a 512 bit vector operations, what you have would be the ability to process four threads of vector operations at a time. This would be absolutely wonderful
though it may be a power usage killer.
The nice thing is that it may not reguire a rewrite at all, just about any multithread app that may have more than one thread doing vector operations would benefit. Along with that the syste would have access to vector operations. Yes heavy single thread vector operations would be slow, but do realize that only that app would be slower any other running software would not be impacted.
Why you would want to elimenate multi threading is beyond me. Multithreading is what makes this sort of machine feasable. Sure a CPU bound process will perform poorly but how much software out there is really that bound up. Lets face it the OS would benefit, just about every application written in the last few years would benefit, people who run multiple programs at anyone time will beenfit. The only people who would loose would be those running CPU bound programs that don;t multi thread well and don't make much use of system resources. Sure there are some programs like this but not many.
Alright who really cares about bench marks anyways, this is an Apple forum right. Really Apple has suffered for years with poorly performing machines, when it comes to bench marks, whats to stop this. Plus any reasonable marketing idiot will come up with multi thread benchmarks that show the potential of this machine. Further if marketed towards resonably intelligent customers 4X SMP will be a big draw in and of itself. It would also be reasonable to assume that bandwidth and VMX issues would be addressed so the 700 MHz performance would exceed the current results that come from the 750 series.
Like I said I have alot of reservations about the reality of this machine. On the other hand the concept is sound and would leverage in a very positive manner some of Apple greatest technology advantages. Sitting here in NY I don't know what Apple has up its sleeves, what I do know is that the sensible thing to do would be to introduce a low power 970 in the powerbook. This would be playing it safe for Apple and would satisfy the lust that their customers have for such a machine. One thing is for certain, Apple is about the only manufacture that could pull this off, that is a massive SMP portable.
Originally posted by Zapchud
Even though Amorph and Wizard69 did a fair job in getting something remotely credible out of this rumor, I still find this rumor to have way too many unanswered questions about it.
How will they get Altivec to perform acceptably?
Like it's been proposed now, we wind up with 4 Altivec-units. This would be pretty equivalent to a very low-clocked, but flexible 512-bit Altivec.
Even 256-bit Altivec has been doomed into the land of the law of diminishing returns, because it will be very hard to find enough parallelism in the code to exploit this (and it would make the unit itself a transistor monster, but that's not the concern here).
Yes, you could probably get acceptable performance out of these four units, if you're skilled enough as a programmer to manage to split the code onto all the units, but I believe that's pretty hard, and more importantly: It would require a rewrite of all code.
On regular code, optimized for the 970, and old (legacy) AV-code, it would perform at ~30% compared to the 970, because of the low clock-frequency, which is pretty low in my book.
This could be possibly be solved with the technology that would have to be the key to the biggest problem of this design entirely: To make it possible for all the cores to execute on the same thread, i.e. eliminate the need for multithreading. (And don't ask me how to do that, if it's even possible)
Without this, this pseudo-G5 Powerbook would be a terrible machine for all things involving less than 4 threads. A typical example of this is benchmarks. Benchmarks are almost always performed with the benchmark itself as the only running CPU-intensive task. And the typical benchmark (especially the cross-platform ones) is not multithreaded.
So unless some breakthrough is achieved relative to multi/single-threading, this machine would probably stand out as very slow in every aspect of benchmarking, except where they tested multitasking extensively.