root@sbcs.UUCP (04/10/87)
> In article <383@sbcs.UUCP> root@sbcs.UUCP (Root) writes: > ... > >My personal loyalty is solely for the fastest (75%), cheapest (5%), > >most standard (20%) silicon at any given point. > > How about architecture? Does that count for anything? > Like I said, I want a *fast* 32 bit environment.. With my application developers hat on, I say "Who cares about architecture??".. > For instance, have you ever tried to use the 80286 in real mode when the > size of your project exceeds 1MB of ROM/RAM. Can you say "bankswitch"? > Do you know what it takes to build a large bankswitched system? I'd > rather watch paint rust than sit at a terminal and link and locate my > code N+1 times (N = number of banks of ROM). And all of this hassle > -- Yes, Clark, I understand all of this (and quite a bit more :-). I helped develop a medical bedside monitor that made extensive use of coroutines, mucho bankswitching, and other yucky contortions all in Pascal on a LSI-11. The point of my previous message is that the average guy writing the next Lotus-123 lookalike in C need not concern himself with machine level monstrousities if the C environment is adequate for his application.. As for choosing a particular chip when your application permits it, I have found that other factors are more important than just instruction set elegance: is there a development system? is a compiler available? is the chip itself available (and if necessary, second sourced)? what is the overall systems hardware cost is using the chip? what is the relative price/performance of different chips used in the context of the system under consideration? what support chips are available for use with the chip? how much mfg support can I expect for the chip? how buggy is the current silicon (thinking of the 32X32 experience here)? > Clark Morgan, Tektronix Lab Instruments Engineering (503) 627-3904 > UUCP: {decvax,cae780,uw-beaver}!tektronix!tekig4!clarkm > US Mail: Tektronix, P.O. Box 500, DS 39-140, Beaverton, OR 97077 Rick