SPICE is still useful software! I remember using it occasionally when doing
my circuits homework back in engineering school (this'll probably get a lot
of groans...) ca. 2002-2003. Much easier for me to figure out than
CADENCE!! :)
Best,
Sean
On Sun, Jun 22, 2014 at 11:00 AM, Paul Koning <paulkoning at comcast.net>
wrote:
On Jun 22, 2014, at 1:26 AM, Mark J. Blair <nf6x at nf6x.net> wrote:
On Jun 21, 2014, at 19:31 , Steven M Jones <classiccmp at crash.com> wrote:
Not sure if this is exactly what you meant, but
SPICE (Simulation
Program with Integrated Circuit Emphasis) springs to mind.
Not exactly what I meant, but very much on-topic. I should see if I can
get SPICE
2G.6 running on my VAX.
There?s an open source SPICE (?ngspice?).
I wonder what sort of software DEC used in their
hardware design and
production back in the early to mid 1980s?
I remember PDP-15 systems being used for PCB CAD around 1978. And in late
1980 my boss, lead DECnet/E designer, left that group a few months after
hiring me as his junior developer, to join a new IC CAD group being put
together at DEC Hudson. I also remember hearing of a thing called DECsim
which sounded like a system (as opposed to IC) simulation tool.
In the late 1980s, the Western Research Lab did some very interesting CAD
work in a project to build a single chip processor (MIPS-like, or
Alpha-like?) in ECL technology. That was unheard of ? no one was doing
VLSI ECL. They developed their own tools that made this possible,
including being able to define a design in a process-independent fashion so
that they could switch from one fab to another ? changing layout constraint
rules ? easily. In the end, I don?t think the chip was built, but it
produced some interesting offshoots such as early designs for heat sinks
and chip test fixtures that would work with chips running at well over 100
watts.
One thing I remember about that CAD suite is that it let you define
subsystems in a number of ways ? geometry, transistor level, logic gate
level, and C code. For example, you could wrap a flip-flop in ?for (i = 0;
i < 64; i++)? to define a 64-bit register. And they could do behavioral
models, power requirement analysis, and chip layout from a single set of
definitions.
I assume at some point there must have been a move to industry tools as
opposed to internally developed ones, but I don?t know when, or to what
extent, that happened.
paul