tim@nmtvax.UUCP (07/31/83)
In a recent article to the net someone compared the T11 and 8088
in the Dec 350 and Rainbow respectively. I'm not sure but I think
one of the instruction timings was off. Were it compares:
T11
ADD R0,(R1) 2.8us
the 8088 timing for this is (5mhz)
ADD BX,(SI) 18cycles or 3.4us not 5.8us
This shows the 8088 is very close in speed for simple operations like
add and subtract. An 8086 would have a wider memory bus and wouldron%brl-bmd@sri-unix.UUCP (08/02/83)
From: Ron Natalie <ron@brl-bmd> Does the DEC 350 contain an 11/23 or a T11? They are not the same. -Ron
BILLW%sri-kl@sri-unix.UUCP (08/02/83)
eed - modern memory is fast.) BillW
tim@nmtvax.UUCP (08/09/83)
In a recent article to the net someone compared the T11 and 8088
in the Dec 350 and Rainbow respectively. I'm not sure but I think
one of the instruction timings was off. Were it compares:
T11
ADD R0,(R1) 2.8us
the 8088 timing for this is (5mhz)
ADD BX,(SI) 18cycles or 3.4us not 5.8us
This shows the 8088 is very close in speed for simple operations like
add and subtract. An 8086 would have a wider memory bus and would
execute the same instruction in 14cycles or 2.8us (they look pretty
even on speed here). My assumption is that you meant add register with
contents of memory addressed by register. I hope thats right. T11
looks like z8000 code and thats what that would mean for it.
Someone else commented that the z80A in the Rainbow was often faster
than the 8088. The z80A isn't really any faster its just that quite a
bit of the 8088 software out there is just tranlated 8080 stuff. This
is really slow and makes very poor use of the 8088. Intel says that you
should get about 20-30% slower speed than a 5mhz 8080 would.
--
Tim Tucker
P.O. Box 3431 C.S.
New Mexico Tech
Socorro, NM 87801
ucbvax!unmvax!nmtvax!tim (uucp)
tim.nmt@rand-relay (arpa)
tim@nmt (CSnet)