🚀 go-pugleaf

jd-wong@uiuc.edu (Jim Wong) wrote:
>In article <415mt0$psf@sundog.tiac.net>, Jim Williams <williams@tiac.net> wrote:
>> I'm not sure if it is my understanding of the term SMP is off, your
>> definition of Hydra is off, or if you are right.  What does SMP mean
>> -- in particular the "S" part?  Also, Exactly what was (is?) Hydra?
>> (I have some **extrememly** vague memory of the word.  I have the
>> feeling that *all* I knew was that there was something called Hydra.)
>> 
>> My gut feeling is that Hydra was an early multi-processor project, but
>> that the methodology was not very similar to what is now called SMP.
>> I admit that I'm guessing.
>
>The C.mmp was a research project at CMU consisting of 16 PDP-11s connected
>through a crossbar to 16 memory modules; the result was a shared-memory
>MIMD system.  Hydra, the operating system of the C.mmp, allowed multiple
>processors to access operating system functionality simultaneously, as
>long they didn't make use of the same data structures.
>

Thanks!  I assume it was ok to access the same data structures as long
as nobody was modifying them.

Now for the SMP part.  "S" is symmetric, right?  What does this mean?
(I think the last thing I read on multi-processors was Hillis's book on
the processor which formed the foundation of Thinking Machines.)

-- 
Sphere.

Find a Linux/GNU Group for you: http://www.tiac.net/users/williams/lugnuts/

Buy Free UNIX!