27 Jan
1999
27 Jan
'99
4:23 a.m.
At 10:21 PM 1/26/99 -0800, Sam Ismail wrote:
To demonstrate that old
hardware that can be picked up for pennies can be combined to attain
amazing amounts of computing power.
Amazing? How many orders of magnitude difference in horsepower
between a C-64 and a $600 Best Buy Intel box? Perhaps I'm being
unromantic, and I certainly have too many old computers of the
XT/AT/486 variety, but I just don't "get" many of these distributed
computing projects. (I do leave my spare contemporary computers
working for the RC5 project at distributed.net, but that's another
story.) Take 10 computers at 1 horse each, and they're still not
equal to one contemporary (cheap) computer at 10 horses. Sure,
there's hack value in doing it, but mostly for people with too
much time on their hands, or for people who aren't paying for the
electric bill or the room to put them in. Hack away, sure - but
claim they're doing "useful" work?
- John
27 Jan
27 Jan
5:20 a.m.
On Wed, 27 Jan 1999, John Foust wrote:
At 10:21 PM 1/26/99 -0800, Sam Ismail wrote:
Yes - and No 8-)
To demonstrate that old
hardware that can be picked up for pennies can be combined to attain
amazing amounts of computing power.
Amazing? How many orders of magnitude difference in horsepower
between a C-64 and a $600 Best Buy Intel box? Perhaps I'm being
unromantic, and I certainly have too many old computers of the
XT/AT/486 variety, but I just don't "get" many of these distributed
computing projects. (I do leave my spare contemporary computers
working for the RC5 project at distributed.net, but that's another
story.) Take 10 computers at 1 horse each, and they're
still not equal to one
contemporary (cheap) computer at 10 horses. Sure, there's hack value
in doing it, but mostly for people with too much time on their hands,
or for people who aren't paying for the electric bill or the room to
put them in. Hack away, sure - but claim they're doing "useful" work?
Well - I wrote a 6502 emulator/simulator that puts the hardware I/O on
seperate machines. It was all written in Perl. Each 6532 (it was a
KIM-1 8-) ran on a diferent machine. When the address of the 6532 got hit
- it talked TCP/IP to the selected machine (talk about address translation
8-) and received back the ROM if it was in the ROM area or performed IO if
it was in the digital IO section. Now you can ask the BIG question of
Why?
It lets me re-write the individual pieces of IO code and replace hardware
at 'anytime'. Handy if you are 'hacking' and it taught me a lot. I
could
change the 'hardware' on the fly and even replace the 6532's with any
other IC I chose. Plus - the code is expandable to include just about
any CPU and any architecture.
No - it is not a speed demon 8-) It does run asynchronously but then so
do most processors. But I can see things change on the fly at each
machine and follow all the registers in all the hardware at about the same
time. I can single step and even micro-step the emulator.
BC
5:50 a.m.
A non-text attachment was scrubbed...
Name: not available
Type: text
Size: 688 bytes
Desc: not available
Url :
http://www.classiccmp.org/pipermail/cctalk/attachments/19990127/e2ae85bd/at…
6:09 a.m.
New subject: Massive parelell old junk
>To demonstrate that old
>hardware that can be picked up for pennies can be combined to attain
>amazing amounts of computing power.
Amazing? How many orders of magnitude difference in
horsepower
between a C-64 and a $600 Best Buy Intel box? Perhaps I'm being
unromantic, and I certainly have too many old computers of the
XT/AT/486 variety, but I just don't "get" many of these distributed
computing projects. (I do leave my spare contemporary computers
working for the RC5 project at distributed.net, but that's another
story.) Take 10 computers at 1 horse each, and they're still not
equal to one contemporary (cheap) computer at 10 horses. Sure,
there's hack value in doing it, but mostly for people with too
much time on their hands, or for people who aren't paying for the
electric bill or the room to put them in. Hack away, sure - but
claim they're doing "useful" work?
So, lets do RC5 un the C64 ?
Of corse, the C64 thing will only get some result on tasks that
are paralell from nature, and of course the result will be more
no 'personal' usable thing. An for shure an pentium box will out-
perform it easy, but as you said, there is a hack value, and see
it from a 1984 perspective, it _will_ be a huge power in that
seting. A cluster of 64 working nodes (plus some 20 communication
nodes) might produce a peek performance of 18 MIPS(6502) - that
can be near a 20 MHz 386 (or 486) - and a 20 MHz 386 was _way_
out of reach at that time.
In fact, I would love to see this thing cracking codes - just to
show that it might have been possible with of the shelf home
computer hardware to crack encodings that the (US) government
wanted us to belive to be safe - 40 Bit or less keys - they still
tell te story - and if such a home computer junk pile can be
a tread, than of course, well designed paralell systems have
been done at tis time ... Just an idea, since I'm realy mad
about 'official' crypto statements.
Gruss
H.
--
Ich denke, also bin ich, also gut
HRK
3:45 p.m.
New subject: Massive parelell old junk
perform it easy, but as you said, there is a hack
value, and see
it from a 1984 perspective, it _will_ be a huge power in that
seting. A cluster of 64 working nodes (plus some 20 communication
nodes) might produce a peek performance of 18 MIPS(6502) - that
can be near a 20 MHz 386 (or 486) - and a 20 MHz 386 was _way_
out of reach at that time.
I've been building a Beowulf cluster out of discarded 486 and Pentium
junk. Even got a 64 port hub to connect the thing. I expect to be below
the 10.9 gflops reported so far by at least an order of magnitude. But
that's hardly the point. That's most why nobody's heard from me on the
list lately. =-D
Anthony Clifton - Wirehead
4:41 p.m.
New subject: Massive parelell old junk
On Wed, 27 Jan 1999, Wirehead Prime wrote:
I've been building a Beowulf cluster out of
discarded 486 and Pentium
junk. Even got a 64 port hub to connect the thing. I expect to be below
the 10.9 gflops reported so far by at least an order of magnitude. But
that's hardly the point.
Ouch. You may be unimpressed by its performance; Beowolf clusters really
want to be switched, over multiple 100 Mbit connections, at that. The
shared fabric of a half-duplex 10 Mbit ethernet at 64 nodes is probably
going to get you the performance of 64 286 machines instead of 64
486-class machines.
Power to you, all the same. It'll be fun at least. (:
--
ok
r. r e d @ b e a r s . o r g
===========================
[ urs longa | vita brevis ]
6:22 p.m.
New subject: Massive parelell old junk
I've been
building a Beowulf cluster out of discarded 486 and Pentium
junk. Even got a 64 port hub to connect the thing. I expect to be below
the 10.9 gflops reported so far by at least an order of magnitude. But
that's hardly the point.
Ouch. You may be unimpressed by its performance; Beowolf clusters really
want to be switched, over multiple 100 Mbit connections, at that. The
shared fabric of a half-duplex 10 Mbit ethernet at 64 nodes is probably
going to get you the performance of 64 286 machines instead of 64
486-class machines. 
9:52 a.m.
On Wed, 27 Jan 1999, John Foust wrote:
At 10:21 PM 1/26/99 -0800, Sam Ismail wrote:
So when this is pulled off successfully and we demonstrate the value of
this project, will you eat a print-out of this e-mail (on a standard 8.5 x
11 piece of paper) in front of the VCF crowd?
Sellam Alternate e-mail: dastar(a)siconic.com
------------------------------------------------------------------------------
Always being hassled by the man.
Coming in 1999: Vintage Computer Festival 3.0
See http://www.vintage.org/vcf for details!
[Last web site update: 01/15/99]
To demonstrate that old
hardware that can be picked up for pennies can be combined to attain
amazing amounts of computing power.
Amazing? How many orders of magnitude difference in horsepower
between a C-64 and a $600 Best Buy Intel box? Perhaps I'm being
unromantic, and I certainly have too many old computers of the
XT/AT/486 variety, but I just don't "get" many of these distributed
computing projects. (I do leave my spare contemporary computers
working for the RC5 project at distributed.net, but that's another
story.) Take 10 computers at 1 horse each, and they're still not
equal to one contemporary (cheap) computer at 10 horses. Sure,
there's hack value in doing it, but mostly for people with too
much time on their hands, or for people who aren't paying for the
electric bill or the room to put them in. Hack away, sure - but
claim they're doing "useful" work?
9313
days inactive
9313
days old
7 comments
7 participants
participants (7)
-
brett＠xnet.com -
ckaiser＠oa.ptloma.edu -
dastar＠ncal.verio.com -
Hans.Franke＠mch20.sbs.de -
jfoust＠threedee.com -
red＠bears.org -
wirehead＠retrocomputing.com