RE: Transputer Development System, 2006? (aka Sony PS/3 runs multi-core Linux)

["Tony Gore" <Tony@xxxxxxxxxxxx>]

I don't doubt that the occam community is thinking of more complex
concurrency.

But, with 4 core processors just arriving, and 8 core processors due
within the next year, and dual processor motherboards around, then this
time next year, a 16 core system is practical, if not common.

By the time you get to this number of cores, shared memory models are
getting tricky and so CSP offers a nice framework of how to think about
the concurrency. We know from the past that there are few problems where
they can be easily "farmed", and there always was a problem with making
general purpose software go faster. The difference 20 years on is that
our software is no longer a single function - we often have many
different and diverse programs running and putting these onto separate
cores does make sense.

Server virtualisation ends up with lots of redundant software running,
and so will always be inefficient. But if major processes or clusters of
processes can be put on a single core and cleanly hooked up to the
"host" OS, then it starts to get more intersting. For example, a web
server process can be replicated to cores, and allows the process to be
kept simpler and more secure.

Take a look at "Sniffing out Microsoft's 'OS in the cloud' skunk-works
project"

http://blogs.zdnet.com/microsoft/index.php?p=119&tag=nl.e589 

Tony Gore

email  tony@xxxxxxxxxxxx 
tel +44-1278-761000  FAX +44-1278-760006  GSM +44-7768-598570
URL: www.aspen.uk.com
Aspen Enterprises Limited
Registered in England and Wales no. 3055963 Reg.Office Aspen House,
Burton Row, Brent Knoll, Somerset TA9 4BW.  UK


-----Original Message-----
From: Barry Cook [mailto:Barry@xxxxxxxxxxxx] 
Sent: 29 November 2006 15:11
To: Tony Gore; Chalmers, Kevin; Matt Jadud; Andrew Delin
Cc: occam-com@xxxxxxxxxx
Subject: Re: Transputer Development System, 2006? (aka Sony PS/3 runs
multi-core Linux)

Hi Tony,

Funnily enough I've just read the IEEE Computer article on
virtualisation (separate virtual OS's) [pp12-14, November issue]. This
is the Nth time virtualisation has crossed my horizon and it appears to
have a lot of support - I guess IBM proved its value back in the VM/370
days and there are now several companies offering software for
single-processor machines.

It seems to me more like having several CPU's just happening to be on
the same silicon (with cost savings in sharing memory etc.) - a rather
limited interpretation of parallel computing. I am certainly finding it
useful to have a "dual" processor so I can leave a compute-intensive
task running and still do email (as now) without it going annoyingly
slowly.

I suspect the occam community is thinking of something more complex in
terms of concurrency.

      Barry.

Dr Barry M. Cook, BSc, PhD, CEng, MBCS, CITP, MIEEE CTO, 4Links Limited,
The Mansion, Bletchley Park,
MK3 6ZP,
UK.
----- Original Message -----
From: "Tony Gore" <Tony@xxxxxxxxxxxx>
To: "Barry Cook" <Barry@xxxxxxxxxxxx>; "Chalmers, Kevin" 
<K.Chalmers@xxxxxxxxxxxx>; "Matt Jadud" <mcj4@xxxxxxxxxx>; "Andrew
Delin" 
<Andrew.Delin@xxxxxxxxxxxxx>
Cc: <occam-com@xxxxxxxxxx>
Sent: Wednesday, November 29, 2006 12:07 PM
Subject: RE: Transputer Development System, 2006? (aka Sony PS/3 runs
multi-core Linux)


Hi Barry

Out in one part of the world, multicore has got some uses - Microsoft's
Virtual Server can allocate a virtual machine to a core, so that it is
possible (say) to have one core allocated to the "host" and three cores
allocated to three separate VMs. Since you can then network these, it is
possible to have very coarse CSP.

Now what would be interesting is if this development were to go further,
and whole chunks of the OS be put onto specific cores.

In my area - I do a lot of Windows Small Business Server support, I
could see a great improvement in security and robustness. For example,
one core handling all the traffic to the outside world e.g. running the
firewall and VPN processes, and everything else communicating through
them would ensure that at least perimeter security could be dealt with
more effectively.

Overall the security and robustness could be improved, because instead
of a monolithic OS that is growing patch by patch, then it is possible
to have a divide and conquer approach that makes use of the multicored
silicon.

Obviously, it would be nice to have a lighter weight inter-process
communication than the full Ethernet stack running on Virtual hardware,
but it seems to me that the server virtualisation is a good target for a
coarse CSP approach, and at this atage, we need to convince people that
CSP is a valid approach.

Tony Gore

email  tony@xxxxxxxxxxxx
tel +44-1278-761000  FAX +44-1278-760006  GSM +44-7768-598570
URL: www.aspen.uk.com
Aspen Enterprises Limited
Registered in England and Wales no. 3055963 Reg.Office Aspen House,
Burton Row, Brent Knoll, Somerset TA9 4BW.  UK




-----Original Message-----
From: owner-occam-com@xxxxxxxxxx [mailto:owner-occam-com@xxxxxxxxxx] On
Behalf Of Barry Cook
Sent: 29 November 2006 11:35
To: Chalmers, Kevin; Matt Jadud; Andrew Delin
Cc: occam-com@xxxxxxxxxx
Subject: Re: Transputer Development System, 2006? (aka Sony PS/3 runs
multi-core Linux)

Kevin,

It probably says nothing about Intel's views on parallel development.

Intel's business is selling chips and if they can offer even slightly
more performance for no more effort from customers (keep the same
software) they will achieve their goal.

Even great gains in performance that include re-development of software
(whether easy or not, anything is more effort than nothing) are more
difficult to sell (recall Inmos).

I'm still trying to find the compelling reason that will convince the
world to jump to parallel software - and thinking along the lines of it
needing to be new software and maybe to reduce power consumption (as is
a paper at Eindhoven). This is leading me to think of embedded systems
that are relatively small (in terms of lines of code as well as for
energy or size).
If you add reliability / safety-critical supported by formalism then
Automotive might be a target (as suggested by Eric Verhulst), as might
Space.

      Barry.

Dr Barry M. Cook, BSc, PhD, CEng, MBCS, CITP, MIEEE CTO, 4Links Limited,
The Mansion, Bletchley Park,
MK3 6ZP,
UK.
----- Original Message -----
From: "Chalmers, Kevin" <K.Chalmers@xxxxxxxxxxxx>
To: "Matt Jadud" <mcj4@xxxxxxxxxx>; "Andrew Delin"
<Andrew.Delin@xxxxxxxxxxxxx>
Cc: <occam-com@xxxxxxxxxx>
Sent: Wednesday, November 29, 2006 10:52 AM
Subject: RE: Transputer Development System, 2006? (aka Sony PS/3 runs
multi-core Linux)


Perhaps a little off topic on this one.  Myself and Jon Kerridge were
looking at Intel's descriptions and plans with multi-core.  From what we
can gather, their plan is to remove any need to develop parallel
systems, and let the system try and work out the best approach.  It
seems Intel believes that parallel development is too difficult.  Sigh.

Kevin Chalmers
Research Student
School of Computing
Napier University
Edinburgh
>
> Hi Andrew,
>
> Damian Dimmich at is working towards this; he had a paper in CPA 2006 
> that explores just this issue, and has a working port of the 
> Transterpreter (a small, portable runtime for occam-pi) to the Cell.
>
>
http://www.transterpreter.org/papers/dimmich-jacobsen-jadud-cpa-2006.pdf
>
> Running on top of Yellow Dog would be the easy way in; Damian is 
> exploring code distribution and code generation for multi-core targets

> like the Cell, and (currently) has 9 separate instances of the runtime

> environment on a single CPU.
>
> See the paper for more details; also, since Damian is on this list, he

> might have additional comments or be able to address more specific 
> questions that you or others might have.
>
> Cheers,
> Matt
>
> Andrew Delin wrote:
> > Team, I thought this was interesting.
> >
> > Why might we be interested in the release of Sony's PS/3 games
console?
> >
> > Because it contains a multi-core Cell processor - and can run Linux.
> >
> > Fred and others - I am wondering if it is possible to release a KROC
> that targets this platform and takes advantage of the multiple
processors
> inside the new Sony console. This would give a true parallel machine
to
> run Occam-Pi. It could be used as a modern 'TDS' with several cores to
run
> on.
> >
> > Nine cores is very tempting - and rather cheap. I understand the YD
> Linux distribution doesn't fully use all cores, but perhaps an
Occam-Pi
> build could? If we can piggy back on the interest in Linux, perhaps we

> might get more interest in the process-oriented-design philosophy
we've
> discussed on this group.
> >

This message is intended for the addressee(s) only and should not be
read, copied or disclosed to anyone else outwith the University without
the permission of the sender.
It is your responsibility to ensure that this message and any
attachments are scanned for viruses or other defects. Napier University
does not accept liability for any loss or damage which may result from
this email or any attachment, or for errors or omissions arising after
it was sent. Email is not a secure medium. Email entering the
University's system is subject to routine monitoring and filtering by
the University.