Pepto-Bismol
Newcomer
MfA said:
Which is why technologists are being so painstaking about it? By the time manufacturing is in full swing, STI will have toiled away at this Frankenstein for the better part of a decade.
Some Innuendo!!
- Thread starter Pepto-Bismol
- Start date
Which is why technologists are being so painstaking about it? By the time manufacturing is in full swing, STI will have toiled away at this Frankenstein for the better part of a decade.
Pepto-Bismol
Newcomer
MfA said:
I think it is a way to ensure that future Cells will be able to emulate past and present ones.
Remember those mnemonics from biology class?
- King Philip Came Over From Greece Sailing-Vessels
- Kittens Prefer Cream Or Fish, Generally Speaking
- Kings Play Cards On Fat Girl's Stomachs
Well, continuing in the same (biological) vein, I think architects will use a similar taxonomy to include or exclude hardware from software. So while different generations of chips (in different products) may all be working together to solve their problems, only particular ones will be able to synthesize what is going on. A TV, for instance, may not be able to understand what PlayStation 3 may be up to, though it could be helping it along ... and vice versa.
Geek Alert: This should make it difficult for incompatible "organisms" to manage (i.e., propagate the data and instructions of) programs that have not been written for them.
It's funny that you should mention DNA. I think the software's "DNA" will force compatible hardware to emulate the organism for which it was written. So that in the future, more potent Cells will carry on as the host "species" would. Hence, the reason for forced inefficiency.
Geek Alert: So if some of you shameful souls daisy-chained a bunch of PS9s together and introduced PS3 code into the system, the organism would emulate PS3.
Panajev2001a
Veteran
DeanoC:
That patent is a bit old: I think it is their pre-CELL research.
That patent is a bit old: I think it is their pre-CELL research
.Pepto-Bismol
Newcomer
MfA said:
Sure. Modern chips may not have the same table manners while gobbling legacy code, but they sure as heck will finish their meal before the earlier chips will! And herein lies the problem. :|
Grid computing is a bit like juggling really. The size of the objects (and how high you toss them into the air) doesn't matter nearly as much as when you expect them to return.
Like most jugglers, you can make adjustments when things come back a little later than planned. The problem, however, is when they come back sooner! And I think this is what the synchronization is for.
Whether the future lies with one particular chip ISA or not, there will always be a battle against the dominant/mainstream ISA of the time, currently the x86. If Cell in the next decade becomes a competitive ISA, I'm sure Intel and Co will try to better it! I remember in the early 90's the ACE/PowerPC consortium for Win NT, tried to abstract MIPS, PowerPC, x86, Alpha etc. ISAs but x86 hardware was the cheapest due to Intel mass production. In the end, Bang for Bucks prevailed...which if Cell lives upto potential could prevail using the PS3 as the trojan horse... 8)
Jaws said:
Try, yes. But they still have to better AMD in many aspects... (And vice versa, admittedly)
