[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [oc] Re: Merlin Hybrid System



Jim,

Converting single thread program/OS to use multiple processors is
a very interesting and long overdue technology leap. Could be an
even bigger leap forward than when we finally dump the x86 instruction
set for a proper RISC optimised set. Have you done any performance
figures yet? Obviously the two things people will look at are
performance and the price. If you could produce a testbench to
demonstrate the theory in practice and obtain some performance
figures I'm sure more companies would be interested.

I think Crusoe have enough on their plate with trying to take on
Intel and AMD in the X86 marketplace. We don't know how successful
their processors have been so they might already be fully stretched
which could help explain their lack of interest.

Paul McFeeters
mailto:paul.mcfeeters@ntlworld.com


-----Original Message-----
From: owner-cores@opencores.org [mailto:owner-cores@opencores.org]On
Behalf Of Jim Dempsey
Sent: 04 December 2001 16:10
To: cores@opencores.org
Subject: Re: [oc] Re: Merlin Hybrid System



----- Original Message -----
From: "Andreas Bombe" <bombe@informatik.tu-muenchen.de>
To: <cores@opencores.org>
Sent: Monday, December 03, 2001 8:02 PM
Subject: [oc] Re: Merlin Hybrid System
...
>
> "It also provides the ability to run multiple operating systems designed
> for the same language to be run simultaneously, enabling two or three
> x86 operating systems to run at the same time." - Virtualization is not
> easy, especially not with systems not designed for it (like x86).  It
> also needs a host OS under which the virtualizer runs.
>

Virtualization of an x86 "language" (instructions?) is commercialy done
now with the Transmeta Crusoe processors. However, Transmeta Crusoe
does not virtualize using multiple processors. Nor does it virtualize
multiple processors using one processor.

I've concieved of a method
by which using an extension of the Crusoe technology that multiple
processors can be used effectively and transparently to make a single
threaded program use multiple threads, make a single processor
operating system use multiple processors, and make an n-processor
operating system use more than n-processors. e.g. when XP professional
is restricted to 2 processors you can use 4, 8, ... This would be done
without any code change to the operating system or applications.

I've done virtualizations in the past. I am the author of some older
operating systems (ETOS and OMNI-8) Virtualization was done at the
operating system level wherein the operating systems that I wrote
provided a virtual machine environments wherein multiple instances
of other operating systems could run. This "older" technology did
require patches to the operating systems that ran on top of the
virtualizing operating system that I wrote. The goal of that product
was to make one machine look like many. OMNI-8 could run on
one or multiple physical processors and provide many more virtual
machine environments running other operating systems (COS310,
OS/8, WPS8).

The new process is an extension to the Crusoe technology and
will not require changes to the operating system. However, note that
the goal of this process will be to make many processors look like one.

I've approached Transmeta with a less-than-white paper on this
concept with no response. I would guess that they are not interested
in expanding their product offerings. I am not sure who I should aproach
to fund the R&D for this. Anyone interested?

Jim Dempsey
President
TAPEDISK Corporation




--
To unsubscribe from cores mailing list please visit
http://www.opencores.org/mailinglists.shtml

--
To unsubscribe from cores mailing list please visit http://www.opencores.org/mailinglists.shtml