HUNEYCUTT@Wpafb-Afwal@sri-unix (12/11/82)
Your assumption was correct...of the current 'popular' crop of 16-bitters, the LSI-11 is slower that all in all benchmarks. (This from 'A Tale of four microprocessors', EDN, April 1, 81.....hmmm, didn't notice the date before) The chips tested were the LSI-11, 8086 (8088 is admittedly a bit slower in memory access), the Z8002, and the 68K. Doug
Seiler@Mit-Xx@sri-unix (12/11/82)
From: Larry Seiler <Seiler@Mit-Xx> The LSI-11 may be slower than the 8088 and other current 16 bit microprocessors, but the DEC Professional PC doesn't have an LSI-11 in it; it uses an F-11, also known (I think) as the 11/23B. I've heard that the F-11 is 1/3 the speed of a PDP-11/70 - anyone know how this compares to the 8088 et al? Oh, and I might also add that J-11 that DEC recently announced (but hasn't started delivering yet) is supposed to be as fast as an 11/70. Now if they made a version of their PC with that inside (and I've heard rumors that they will), that would be one powerful personal computer. Larry -------
ajh (12/13/82)
About the PDP-11/LSI-11/T-11/etc. Here's my two cents worth: The 11's may be slow, but the instruction set is a real joy to use. After using various bizarre machines, I'm always glad to come back to my good old LSI-11. (I'm only talking about 8/16 bit micros.) The reason is that the instruction set is so nicely orthagonal. You say ADD R0,R1 or ADD MEM1,MEM2 or ADD #10,PC. ~vi (Oh well, no editor...) You don't have to load your accumulator or adjust your index registers or do any other really fun stuff. The 11's might not have block moves, 67-bit floating point multiplies, fubar instructions or any other "powerful" feature, but the orthagonal instruction set is easy to use and easy for compiler's, too. (The floating point instructions work very nicely on the stack. The EIS instructions aren't that logical, Oh well...) Anyway, I like PDP/LSI/T-11's (or whatever), Alan J. Hu U.C. San Diego ...ucbvax!sdcsvax!ajh (or something like that)