2 Jun
2017
2 Jun
'17
7:20 a.m.
I was looking at an old GI catalog and casually noting the CP1610 that was most of a PDP11
processor. I did some more web surfing and noticed that the Intellivision game machine
used this chip. It just never dawned on me that they used this processor.
I see that one could even get a keyboard for these.
Dwight
9:26 a.m.
On 06/02/2017 07:55 AM, Al Kossow via cctalk wrote:
A very odd version of the PDP-11
I did some programming on it for GI in the late 70's using their
GIMINI development system and cross-development tools on their Sigma
9.
I sold a bare CP1600 chip about a year ago to a collector. "Odd" is an
understatement. A 10-bit wide instruction word, with the upper 6 bits of
the opcode unused. Loading a 16-bit address took three words.
Also, slow, very slow, with no I/O instructions. A stripped-down
instruction set (IIRC, exclusive OR and AND Boolean ops, but no
inclusive OR). We'd examined its possible usage, but determined that
code would run slower on it than on most contemporary 8-bit CPUs.
But, for 1975, notable because it's a 16-bit monolithic CPU among a
handful of others (Fairchild 9440, MicroNova, National PACE, TI 9900)
--Chuck
9:37 a.m.
I sold a bare CP1600 chip about a year ago to a
collector. "Odd" is an
understatement. A 10-bit wide instruction word, with the upper 6 bits of
the opcode unused. Loading a 16-bit address took three words.
Also, slow, very slow, with no I/O instructions.
But that was because it has memory-mapped I/O, no? On the other hand the
decles were weird and it has a lot of instructions that were removed.
Retrospectively a 6502 or a Z80 would have looked like a better choice in
this application, even considering this was supposed to be a higher-end
console.
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- Knowledge puffs up, but love builds up. -- 1 Corinthians 8:1 ---------------
10:04 a.m.
On 06/02/2017 09:37 AM, Cameron Kaiser wrote:
But that was because it has memory-mapped I/O, no? On
the other hand
the decles were weird and it has a lot of instructions that were
removed.
GI made a 10-bit wide ROM for program storage, as I recall. I think the
10-bit instruction word was the single greatest liability, resulting in
inefficient memory usage when memory was expensive--and slower execution
time.
For the time, not a bad effort, particularly when contrasted with the
National 16-bit PACE, which was loosely based on the 4-accumulator DG
Nova design and which required special buffers for integration into TTL
designs.
--Chuck
11:02 a.m.
Such things were often done on loyalty to previous companies.
It is like the Video Brain used the F8 or that Olivetti used the Z8000 for the M20.
The Video Brain was because the designer had worked on the F8 at Fairchild.
The M20 was because Faggin was Italian and had connections to Olivetti.
Such things were not always done because they were better or cheaper.
Woz was rare in his way of thinking at the time. He was mostly looking for cheap
but having a rich set of addressing mode was surely a plus.
Dwight
________________________________
From: cctalk <cctalk-bounces at classiccmp.org> on behalf of Cameron Kaiser via
cctalk <cctalk at classiccmp.org>
Sent: Friday, June 2, 2017 9:37:13 AM
To: cclist at sydex.com; cctalk at classiccmp.org
Subject: Re: I hadn't made the connection before
I sold a bare CP1600 chip about a year ago to a
collector. "Odd" is an
understatement. A 10-bit wide instruction word, with the upper 6 bits of
the opcode unused. Loading a 16-bit address took three words.
Also, slow, very slow, with no I/O instructions.
But that was because it has memory-mapped I/O, no? On the other hand the
decles were weird and it has a lot of instructions that were removed.
Retrospectively a 6502 or a Z80 would have looked like a better choice in
this application, even considering this was supposed to be a higher-end
console.
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com<http://www.floodgap.com> *
ckaiser at floodgap.com
-- Knowledge puffs up, but love builds up. -- 1 Corinthians 8:1 ---------------
9:48 a.m.
On 6/2/2017 11:26 AM, Chuck Guzis via cctalk wrote:
On 06/02/2017 07:55 AM, Al Kossow via cctalk wrote:
This begs the
question of why the CP1600 was designed that way. I fail
to believe that GI engineers were somehow less intelligent than the rest
of the population, so what happened? Upper Management dictate?
Customer demand? It would be interesting to know.
Jim
A very odd version of the PDP-11
I did some programming on it for GI in the late 70's using their
GIMINI development system and cross-development tools on their Sigma
9.
I sold a bare CP1600 chip about a year ago to a collector. "Odd" is
an
understatement. A 10-bit wide instruction word, with the upper 6 bits of
the opcode unused. Loading a 16-bit address took three words.
Also, slow, very slow, with no I/O instructions. A stripped-down
instruction set (IIRC, exclusive OR and AND Boolean ops, but no
inclusive OR). We'd examined its possible usage, but determined that
code would run slower on it than on most contemporary 8-bit CPUs.
But, for 1975, notable because it's a 16-bit monolithic CPU among a
handful of others (Fairchild 9440, MicroNova, National PACE, TI 9900)
9:13 a.m.
I was looking at an old GI catalog and casually noting
the CP1610 that was
most of a PDP11 processor. I did some more web surfing and noticed that the
Intellivision game machine used this chip. It just never dawned on me that
they used this processor.
I see that one could even get a keyboard for these.
The (ridiculous) story of the Keyboard Component was legendary. The ECS
keyboard variant can barely be considered functional even by the standards
of the time, though I guess it at least looks decent compared to an Aquarius
and the second sound chip was nice. It was designed to be cheap and get the
FTC off Mattel's back and that's all it did.
I say this as a kid whose first video game system was, in fact, an Inty.
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- He whose face gives no light, shall never become a star. -- William Blake --
2525
days inactive
2525
days old
7 comments
5 participants
participants (5)
-
aek＠bitsavers.org -
brain＠jbrain.com -
cclist＠sydex.com -
dkelvey＠hotmail.com -
spectre＠floodgap.com