dual processors and linux 1768
Actually, the original loop was mostly CPU bound, because essentially all the disk access were very quickly cached.
Roy L. Fuchs I think they took the time to make an SMP kernel because it is easy enough to say "just use a double (or quadruple) speed processor", but not all...
The gzip command above is running while I am typing this response. 'top' shows CPU use at 97-98% on most snapshots, except when I do something else interactively on the system.
My posting was made in response to the following:
In the middle of composing this, on my single-CPU system with no hyperthreading, while the gzip command continued to run, I did several text captures from Firefox windows, brought up and read several news stories using Firefox, brought up a root terminal window and ran chkrootkit in it. While chkrootkit ran, system interactive response was a little slower than normal. Without chkrootkit running, with only gzip running, system interactive response was only a very tiny bit less snappy than with nothing else heavy running.
That is almost certainly the case. (However, my test case used Mandriva Linux 2006, not a Mac.) I recall being informed in either around 1987 and maybe around 1980 that Unix adjusted a process's...
The person who claimed "any of todays preemptive mulbreastasking OS will ... slow to a craw on a single CPU" was clearly mistaken. That was and is my point.
Yes, that is what I meant by the word "niced" in the next to last line of my paragraph to which you responded. Using 'nice' on a CPU-bound job does eliminate almost all of even the tiny bit of slowness (to interactive response) a CPU-bound job will cause on a single-CPU system.
I will agree with the multi-CPU fans that a SMP system will show much less slowing (if any is detectable) for a given load than a comparable single-CPU system. However, a single-CPU system won't "slow to a craw sic" as had been claimed.
-- Robert Riches (Yes, that is one of my email addresses.)