24 Oct
2005
24 Oct
'05
5:20 a.m.
On Sun, 23 Oct 2005, Sridhar Ayengar <ploopster at gmail.com> wrote:
Johnny Billquist wrote:
There basically isn't any (hardware, that is).
The 11/782 was not an SMP design. It was two CPUs with shared memory,
where one CPU acted as a slave to the other. All I/O was on the primary
CPU as well. The second CPU was just a computing resource.
I haven't (yet) found any references to any interconnect hardware apart
But the hardware I still keep around is much more
interesting... :-)
Sometime in the future, I wouldn't mind trying to put an 11/782
together. Anyone know where I might be able to find some information
about the interconnect hardware? from the dual-ported memory boxes.
Maybe someone else knows some more details?
Appearantly, it wasn't much faster than a normal 11/780 either, so some of
the machines appearantly got split up into two separate 11/780s instead.
Johnny
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at update.uu.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
24 Oct
24 Oct
6:58 a.m.
On 24/10/2005, at 7:20 PM, Johnny Billquist wrote:
There basically isn't any (hardware, that is).
The 11/782 was not an SMP design. It was two CPUs with shared memory,
where one CPU acted as a slave to the other. All I/O was on the
primary
CPU as well. The second CPU was just a computing resource.
I haven't (yet) found any references to any interconnect hardware
apart
from the dual-ported memory boxes.
As far as I can recall, that's all you need (apart from a couple of
working 11/780s). The shared memory cabinet contained 4Mb of multi-
port memory. Two were used for the 11/782 but I think there were four
ports altogether. MA780 is the part number that springs to mind but
MA782 makes more sense :-)
Maybe someone else knows some more details?
Treat all comments as being restored from a nearly 20 year old 1/2"
tape with no error checking or correction.
Appearantly, it wasn't much faster than a normal
11/780 either, so
some of
the machines appearantly got split up into two separate 11/780s
instead.
Certainly the only 11/782 I know of (Melbourne University) was split
to get much better performance. Basically all I/O (and hence all of
VMS) had to occur in the "master" 11/780 so all that could be run in
the slave was CPU intensive jobs. As soon as your CPU bound process
wanted to do something other than compute it had to be scheduled in
the "master".
Effectively you had one big global lock on VMS. The story goes that
all the necessary VMS changes to support the 11/782 occurred over a
weekend. Another interesting story concerns the mythical 11/787 (ie
two 11/785s rather than 11/780s). There were rumors of such systems
occurring in a mythical site somewhere near the centre of Australia :-)
Huw Davies | e-mail: Huw.Davies at kerberos.davies.net.au
Melbourne | "If soccer was meant to be played in the
Australia | air, the sky would be painted green"
1 p.m.
Huw Davies wrote:
Certainly the only 11/782 I know of (Melbourne
University) was split
to get much better performance. Basically all I/O (and hence all of
VMS) had to occur in the "master" 11/780 so all that could be run in
the slave was CPU intensive jobs. As soon as your CPU bound process
wanted to do something other than compute it had to be scheduled in
the "master".
The 782 certainly was two 780s joined at the hip. I thought that the
SBI was joined up too but that might be the Purdue (?) VAX-11/784.
When proper SMP was introduced in VAX/VMS V5.0 (the previous support
for the VAX-11/782 being dubbed ASMP), the 782 support was dropped
and the choices were either don't upgrade or split into two 780s.
Effectively you had one big global lock on VMS. The
story goes that
all the necessary VMS changes to support the 11/782 occurred over a
weekend. Another interesting story concerns the mythical 11/787 (ie
two 11/785s rather than 11/780s). There were rumors of such systems
occurring in a mythical site somewhere near the centre of Australia
There would have been no reason that a "787" would not have worked,
but given that bad rep that the 782 had, why would anyone have bothered?
Interestingly I don't think I've ever come across any technical docs
that describe the 782. It does rate a passing mention in the VIDSM.
The various "how we wrote support for the 782" articles that I've read
fail to mention that it was done over a weekend :-)
Somewhere or other I have Kathleen Morse's "how we might do
[A]SMP in VMS" memo. I think the memo itself was long enough
to take a weekend to write!
Antonio
--
Antonio carlini
arcarlini at iee.org
3:45 p.m.
On Monday 24 October 2005 12:00, a.carlini at ntlworld.com wrote:
Huw Davies wrote:
Purdue is correct, 11/784 isn't. The 11/784 is the quad-CPU version of
the 11/782. Purdue's dual-VAX didn't have a DEC model number, because
they never (officially) sold it.
Sharing the SBI made a much more usable dual-processor machine, where
both CPUs could access I/O and *all* of the available memory, not just
4MB of it.
Pat
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
Certainly the only 11/782 I know of (Melbourne
University) was
split to get much better performance. Basically all I/O (and hence
all of VMS) had to occur in the "master" 11/780 so all that could
be run in the slave was CPU intensive jobs. As soon as your CPU
bound process wanted to do something other than compute it had to
be scheduled in the "master".
The 782 certainly was two 780s joined at the hip. I thought that the
SBI was joined up too but that might be the Purdue (?) VAX-11/784.
4:17 p.m.
Patrick Finnegan wrote:
Purdue is correct, 11/784 isn't. The 11/784 is
the quad-CPU version
of the 11/782. Purdue's dual-VAX didn't have a DEC model number,
because they never (officially) sold it.
I remember reading the paper describing Purdue's VAX and it was
definitely home-brew in the sense that they bought bits and
built something. So I'm not surprised that DEC never sold it.
I can find no evidence that DEC ever sold a VAX-11/784 (or any
quad processor VAX-11/780). I don't doubt that someone might
have cobbled something together in a lab on a rainy day
(there were enough DEC research labs gluing stuff together to
see what happened!), but I see no evidence that anything
like that was ever done for a customer.
All the stuff I see on the net is just "not many people know..."
type of stuff with no credible sources (actually, mostly no
sources whatsoever). I know that there were "VAXen that never
shipped" and I might be able to remember the names of some of
them if pushed (Raven, for one). But I never heard even a
whispering of such a VAX while inside DEC and there is nothing
that says "784" in the Options & Modules List (except for
a raft of power supplies, some Mxxx[x] boards and some
DECmate bits). Admittedly the only non-scanned searchable-text
one I have is from 1994, but it still has the VAX-11/780 and
a boatload of PDP-11s in there, so it would be odd if the
mythical VAX-11/784 dropped off by accident.
So unless someone digs out some evidence, I'll just choose
not to believe that DEC ever built a 784 (and after building
the 782, I cannot believe that they would have even tried :-)).
One of the VAX handbooks does have a section in the back that
suggests using butterfly configs of VAX-11/780s with dual-ported
memory to perform FFTs and the like. No suggestion that such
a config would be sold as anything other than individual VAXen
and associated memory boxes. I do know that some fool somewhere
went and gave names to groups of clustered nautilus VAXes and
(IIRC) someone else went and repeated the mistake for at least
one config of VAXft machines, but at least they had the decency
to produce glossy brochures so we could see how confused they
were trying to make us :-)
Antonio
--
Antonio carlini
arcarlini at iee.org
9:14 p.m.
New subject: 11/782
a.carlini at ntlworld.com wrote:
All the stuff I see on the net is just "not many
people know..."
type of stuff with no credible sources (actually, mostly no
sources whatsoever). I know that there were "VAXen that never
shipped" and I might be able to remember the names of some of
them if pushed (Raven, for one). But I never heard even a
whispering of such a VAX while inside DEC and there is nothing
that says "784" in the Options & Modules List (except for
a raft of power supplies, some Mxxx[x] boards and some
DECmate bits). Admittedly the only non-scanned searchable-text
one I have is from 1994, but it still has the VAX-11/780 and
a boatload of PDP-11s in there, so it would be odd if the
mythical VAX-11/784 dropped off by accident.
So unless someone digs out some evidence, I'll just choose
not to believe that DEC ever built a 784 (and after building
the 782, I cannot believe that they would have even tried :-)).
Antonio
What was the name of the dual 11/785 (made by upgrading two halves of
the 11/782 to the 11/785...
I seem to remember seeing 11/785-2 stickers in some upgrades to Field
Service. (These were for non-fcc upgrades of 11/780's to 11/785's... IIRC)
Was the dual 11/785 given the 11/784 name -- or was it an 11/787 or
something like that.
Too damned long since I left 78A/79J Field Service at DEC.
Bill
25 Oct
25 Oct
11:29 a.m.
New subject: 11/782
Bill Pechter wrote:
a.carlini at ntlworld.com wrote:
Everything I've ever seen has referred to it as the 11/787.
Peace... Sridhar
All the stuff I see on the net is just "not
many people know..."
type of stuff with no credible sources (actually, mostly no
sources whatsoever). I know that there were "VAXen that never
shipped" and I might be able to remember the names of some of
them if pushed (Raven, for one). But I never heard even a
whispering of such a VAX while inside DEC and there is nothing
that says "784" in the Options & Modules List (except for
a raft of power supplies, some Mxxx[x] boards and some
DECmate bits). Admittedly the only non-scanned searchable-text
one I have is from 1994, but it still has the VAX-11/780 and
a boatload of PDP-11s in there, so it would be odd if the
mythical VAX-11/784 dropped off by accident.
So unless someone digs out some evidence, I'll just choose
not to believe that DEC ever built a 784 (and after building
the 782, I cannot believe that they would have even tried :-)).
Antonio
What was the name of the dual 11/785 (made by upgrading two halves of
the 11/782 to the 11/785...
I seem to remember seeing 11/785-2 stickers in some upgrades to Field
Service. (These were for non-fcc upgrades of 11/780's to 11/785's... IIRC)
Was the dual 11/785 given the 11/784 name -- or was it an 11/787 or
something like that.
Too damned long since I left 78A/79J Field Service at DEC.
6851
days inactive
6852
days old
6 comments
6 participants
participants (6)
-
arcarlini＠iee.org -
bqt＠Update.UU.SE -
huw.davies＠kerberos.davies.net.au -
pat＠computer-refuge.org -
pechter＠gmail.com -
ploopster＠gmail.com