cruess@oakhill.UUCP (Michael Cruess) (03/04/85)
>From: david@daisy.UUCP (David Schachter) >Chip designers have to make many trade-offs. It appears that National chose >speed-to-market and manufacturability.. Motorola chose prettiness. Time will >tell whose strategic planning department chose better, three or four years ago. It may well be that National chose 24 bits because of speed-to-market and manufacturability. Those were factors in the choice of 24 address pins on the MC68000 several years ago. For the MC68020, they were not. The MC68020 already had 32 bit ALU's, the die is not bonding pad limited in size, and 8 fewer pins would not have put the part in a different package. To restate in a more positive fashion: the MC68020 would not have been smaller, sooner or cheaper if it only had 24 address pins. What we did choose was not having to do it over again. >Mr. Cruess (of Motorola) states that, in his experience, 24 bit addresses are >a bit squeezed and that 28 bits is better. Therefore, he concludes, the >MC68020 is better than the NS32032. In my experience, 24 bits is big enough >to hold us workstation vendors until late 1987 or 1988. Until then, given >the real-world capabilities of microprocessors (as opposed to the claims of >the marketing departments of the microprocessor vendors), anyone worrying >about putting >16 MB on a microprocessor is wasting time. You wouldn't >put a 300 gallon gas tank in a Honda Civic, would you? It would be overkill. In other words, you will not be needing more than 24 bits of addressing for 3 years. Considering how much flak all microprocessor vendors take about how long it takes to bring a product into full production, you should be glad that we are starting now. Also, re: ">16 MB": I still maintain that in discussing microprocessor address pins we are discussing logical (virtual) addresses (as opposed to physical (real) addresses), and that logical spaces of more that 2^24 bytes are required by our customers. The amount of physical memory is not relevant. Please see other articles in this group. >By the time you need >16 MB in a workstation, National Semi will either have >a part to handle your needs or be bankrupt. Until then, the NS32032 gives you >what you want (as does the Intel 80286) without paying for extra, unneeded >address pins and without wasting expensive printed circuit board real estate, >tranceivers, etc. Bankrupt? Do you know something that I don't? As for the rest, see responses above. Michael Cruess Motorola Microprocessor Products Division {ihnp4,seismo,ctvax,gatech}!ut-sally!oakhill!cruess
david@daisy.UUCP (David Schachter) (03/07/85)
In article 350@oakhill.UUCP, Michael Cruess of the Motorola Microprocessor Products division responds to an earlier posting of mine thusly: >In other words, you will not be needing more than 24 bits of addressing for >3 years. Considering how much flak all microprocessor vendors take about >how long it takes to bring a product into full production, you should be >glad that we are starting now. Also, re: ">16 MB": I still maintain that >in discussing microprocessor address pins we are discussing logical >(virtual) addresses (as opposed to physical (real) addresses), and that >logical spaces of more that 2^24 bytes are required by our customers. The >amount of physical memory is not relevant. > 1) Intel managed to get its 80286 into full production a year after delivering working samples to customers. Surely Motorola won't take three years for the 68020, will it? (*-) 2) How many chips in the 80286 class or above are going into sockets in which address spaces >16MB are required? Less than 1/10 of 1 percent? Until large address spaces are routinely needed in the microworld, (about three years for us workstation vendors), 32 bit addressing is overkill. 3) Whether virtual or physical is immaterial. We don't need 32 bit addressing yet (in 99.9 % of the sockets.) We don't need it in the virtual space or in the physical space. We will in three years or so. By then, Moto, Natty Semi, and even Intel will have 32 bit machines. (Intel's 80386 is due for customer sampling in the first quarter of 1986, as of three weeks ago.)
david@daisy.UUCP (David Schachter) (03/07/85)
Oh, I forgot on my previous posting:
[generic disclaimer] {N.F.Q.}
Sorry for the inconvenience. Legalities, you know.