xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) (06/27/90)
In article <22863@snow-white.udel.EDU> EVERHART@arisia.dnet.ge.com writes: >I had the pleasure awhile back of talking with one of the Mach >developers. The gist of the conversation was that the Mach effort >is intended to result in code which can be given out everywhere [...] Bravo! >It's a serious issue for those of us no longer at a college/university. It's been a long time since I looked at the source license cost for a non-educational AT&T Unix installation, but the figure $30,000 sticks in my mind (and craw). That much money is a serious issue for anyone. > An implication of this is that the CBM Unix for Amiga will not >make getting Mach up on Amiga a trivial job. Should someone get the >basic functions running with an OSF or a Gnu kernel, though, >perhaps CBM will see the wisdom of support for this with their >libraries. Wouldn't it be nice? Pretty tough to do when they've already licensed the AT&T code, though. Essentially I'd guess that eliminates all current CBM employees from development on a PD Mach release, due to non-disclosure contract clauses with AT&T. >Having an OS engine under the hood that will allow use >of multiprocessor 68040's could turn out to be a technically simpler >way to progress than throwing out the entire 680x0 code base for >a new engine. Yes! I have no idea of the Mach internals; if it is a "supervisor and subordinates" paradigm, then lots of cheap coprocessors on a plug in board might be doable. If it is a "many peers" paradigm, then boardfulls of 68040's could run less tightly coupled. Either way, talk about your screaming graphics! Where's that tower with all the slots!?! Kent, the man from xanth. <xanthian@Zorch.SF-Bay.ORG> <xanthian@well.sf.ca.us> -- in the distance a roasted cave newt screamed in agony -- Andrew Palfreyman
jerbil@chamber.caltech.edu (Joseph R. Beckenbach) (06/28/90)
xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) writes: >Yes! I have no idea of the Mach internals; if it is a "supervisor and >subordinates" paradigm, then lots of cheap coprocessors on a plug in >board might be doable. If it is a "many peers" paradigm, then boardfulls >of 68040's could run less tightly coupled. Either way, talk about your >screaming graphics! Where's that tower with all the slots!?! There are three ways to organize parallel processing hardware: a processor shares memory with other processors (global memory), a processor does not share memory with other processors (distributed memory), and a hybrid of the two. (For a small number of processors (<16) the global memory scheme is okay; about 128 processors, and distributed-memory starts looking really good. The distributed-memory scheme scales quite easily through thousands.) Mach assumes a shared-memory machine, and can tolerate being run on a hybrid machine. Local-memory machines, such as anything based on the Connection Machine, or on Caltech's Cosmic Cube (Intel ipsc's, Symult 2010), or on transputers [:-b] would be able to run Mach; however, they would each need to run on top of a global-memory simulator to supply the conditions for which Mach is suited. This is a major performance and parallelism loss for the distributed-memory machine. Mach on the Amiga would work well for however many gangs of processors you wish to cluster onto a given memory pool. In fact, a collection of four processors and shared memory on a board, which can use the Amiga buses as a network, would work well. I personally am much more interested in the distributed-memory cases, since I want to eventually use an Amiga as an interface into a box with 1024 (or so :-) processors. Lots of memory, lots of computing cycles. Lots of power :-). Ah well, back to the coal mines. Joe Beckenbach Caltech CS department jerbil@csvax.caltech.edu joe@vlsi.caltech.edu
matt@sapphire.jpl.nasa.gov (matt of ASTD) (06/29/90)
In article <jerbil.646590818@chamber> jerbil@chamber.caltech.edu (Joseph R. Beckenbach) writes: [stuff deleted] > Mach assumes a shared-memory machine, and can tolerate being run >on a hybrid machine. Local-memory machines, such as anything based on >the Connection Machine, or on Caltech's Cosmic Cube (Intel ipsc's, Symult 2010), How does Mach run on a SISD machine like the Connection Machine? I'll can see how it could use Hypercubes or ``Cosmic Cube'', but SISD. Is there some documentation on this. If so, please tell me, I would like to see it. ---------------------------------------------------------------------- Matthew Presley (UCLA CS Grad. Student) & (JPL CS dude) Internet (presley@cs.ucla.edu) or (matt@sapphire.jpl.nasa.gov) "Twisted yellow puppies play broken flutes loudly..."
matt@sapphire.jpl.nasa.gov (matt of ASTD) (06/29/90)
Oops. I got carried away with my Sssssss's. I mean SIMD. With an M. Not SISD (that would be kind of boring). SIMD, SIMD. I always do this. ---------------------------------------------------------------------- Matthew Presley (UCLA CS Grad. Student) & (JPL CS dude) Internet (presley@cs.ucla.edu) or (matt@sapphire.jpl.nasa.gov) "Twisted yellow puppies play broken flutes loudly..."