Sign in

username:

password:



Not a member?

Search fpga-cpu



Search tips

Subscribe to fpga-cpu



fpga-cpu by Keywords

Altera | CISCifying | IDE | ISA | Java | JHDL | JTAG | LBU | MicroBlaze | PAR | PCI | RISC | SoC | Spartan | Transputers | Verilog | VHDL | Virtex | VLIW | WebPack | Xilinx | Xsoc | YARD-1A

Ads

Discussion Groups

Discussion Groups | FPGA-CPU | A bit of history

This list is for discussion of the design and implementation of field-programmable gate array based processors and integrated systems. It is also for discussion and community support of the XSOC Project (see http://www.fpgacpu.org/xsoc).

A bit of history - veronica_merryfield - Feb 10 15:40:00 2005


Mostly. This is neat

http://news.com.com/Hobbyist+reconstructs+Apollos+computer/2100-
1003_3-5570963.html?tag=nefd.top

and so are some of the links from http://starfish.osfn.org/AGCreplica/

There is some real interesting stuff to read on the instruction set
and the inner workings of the CPU and don't forget when this was done.

Veronica




(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )


Re: A bit of history - rtstofer - Feb 10 17:34:00 2005



Now that is a retro project! I am almost tempted to push everything
off my table and stick that thing in a Spartan 3 Starter board with
PS/2 input and VGA output.

What a great way to spend quality time on engineering!

--- In , "veronica_merryfield"
<veronica.merryfield@t...> wrote:
>
> Mostly. This is neat
>
> http://news.com.com/Hobbyist+reconstructs+Apollos+computer/2100-
> 1003_3-5570963.html?tag=nefd.top
>
> and so are some of the links from
http://starfish.osfn.org/AGCreplica/
>
> There is some real interesting stuff to read on the instruction
set
> and the inner workings of the CPU and don't forget when this was
done.
>
> Veronica





(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: A bit of history - Rob Finch - Feb 11 1:05:00 2005


> Mostly. This is neat
>
> http://news.com.com/Hobbyist+reconstructs+Apollos+computer/2100-
> 1003_3-5570963.html?tag=nefd.top
>
> and so are some of the links from
http://starfish.osfn.org/AGCreplica/
>
> There is some real interesting stuff to read on the instruction set
> and the inner workings of the CPU and don't forget when this was
done. Wow! That's a really cool project. I've been wondering what goes into
a guidance computer. Now I can build my own spacecraft :)

Rob



(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: A bit of history - Asif Iqbal - Feb 11 1:24:00 2005


thats a very good link and its very intresting to see
what goes inside a spacecraft... that too in 1964
good job
--- veronica_merryfield
<> wrote: >
> Mostly. This is neat http://news.com.com/Hobbyist+reconstructs+Apollos+computer/2100-
> 1003_3-5570963.html?tag=nefd.top
>
> and so are some of the links from
> http://starfish.osfn.org/AGCreplica/
>
> There is some real interesting stuff to read on the
> instruction set
> and the inner workings of the CPU and don't forget
> when this was done.
>
> Veronica >
>
> To post a message, send it to:
>
> To unsubscribe, send a blank message to:
>
> Yahoo! Groups Links

__________________________________




(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: Re: A bit of history - John Kent - Feb 11 2:51:00 2005

There used to be a car advertisment on TV (SAAB or BMW or something)
that claimed their car had more computing power than that required to
put a man on the moon. Engineers had a bit of a chuckle at that, because it
was known that the guidance computer was not much more powerful than a
Z80 ;-)

Sounds like amazing dedication.

John.

Rob Finch wrote: >Wow! That's a really cool project. I've been wondering what goes into
>a guidance computer. Now I can build my own spacecraft :)
>
>Rob >

--
http://members.optushome.com.au/jekent



(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: A bit of history - rtstofer - Feb 11 3:21:00 2005



The AGC has a truly bizarre architecture - clearly application
specific. Because the machine was so limited in memory they
actually wrote and interpreter and most application programs were
interpreted strictly to save memory. They could use lots of lights -
the machine was slow enough to see them flash!

The AGC used a 1.024 MHz clock and the fast instructions used 12
clock cycles. Memory cycles were 11.7 uS. The 8080 started out at
2 MHz, fast instructions used 2 clock cycles, memory access was 1 uS
(IIRC) and the address space was much larger. --- In , John Kent <jekent@o...> wrote:
> There used to be a car advertisment on TV (SAAB or BMW or
something)
> that claimed their car had more computing power than that required
to
> put a man on the moon. Engineers had a bit of a chuckle at that,
because it
> was known that the guidance computer was not much more powerful
than a
> Z80 ;-)
>
> Sounds like amazing dedication.
>
> John.
>
> Rob Finch wrote:
>
> >
> >Wow! That's a really cool project. I've been wondering what goes
into
> >a guidance computer. Now I can build my own spacecraft :)
> >
> >Rob
> >
> >
> >
>
> --
> http://members.optushome.com.au/jekent




Re: A bit of history
Re: A bit of history


(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: Re: A bit of history - John Kent - Feb 11 3:32:00 2005

Hi Richard,

The 6800 I understand because of one (E) Clock = 1 cycle = 1usec.

Anyway ... what are you doing up this time of night ?
6pm in Australia ... not sure what that is in the US,
but I bet it's early in the morning :-)

John.

rtstofer wrote: >The AGC has a truly bizarre architecture - clearly application
>specific. Because the machine was so limited in memory they
>actually wrote and interpreter and most application programs were
>interpreted strictly to save memory. They could use lots of lights -
> the machine was slow enough to see them flash!
>
>The AGC used a 1.024 MHz clock and the fast instructions used 12
>clock cycles. Memory cycles were 11.7 uS. The 8080 started out at
>2 MHz, fast instructions used 2 clock cycles, memory access was 1 uS
>(IIRC) and the address space was much larger. >

--
http://members.optushome.com.au/jekent



(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: Re: A bit of history - Eric Smith - Feb 11 4:00:00 2005

> The 8080 started out at 2 MHz,

So far so good. Later the 8080A-2 was rated at 2.5 MHz,and the 8080A-1
at 3 MHz.

> fast instructions used 2 clock cycles,

No. Instructions take between one and five "M cycles", each of which
can perform one memory cycle and/or some internal operations. Each
M cycle takes between one and five "T states" (clock cycles). The
fastest instructions take four clocks, and the slowest (XTHL) takes 18
clocks.

> memory access was 1 uS

The memory access time is misleading. Memory cycle time is more useful.
For a 2 MHz 8080, an instruction fetch cycle takes 4 cycles (2 us), and
other memory cycles take at least 3 cycles (1.5 us).

Reference: Intel 8080 Microcomputer Systems User's Manual, September 1975,
Intel literature part number 98-153C.



(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: A bit of history - rtstofer - Feb 11 11:31:00 2005



You are correct, of course. I had misplaced my 8080 Microcomputer
Systems User's Manual (Sept 1975). I have since located it - right
on top of my "Bit-Slice Microprocessor Design" book. I just have to
get organized someday... --- In , "Eric Smith" <eric@b...> wrote:
> > The 8080 started out at 2 MHz,
>
> So far so good. Later the 8080A-2 was rated at 2.5 MHz,and the
8080A-1
> at 3 MHz.
>
> > fast instructions used 2 clock cycles,
>
> No. Instructions take between one and five "M cycles", each of
which
> can perform one memory cycle and/or some internal operations. Each
> M cycle takes between one and five "T states" (clock cycles). The
> fastest instructions take four clocks, and the slowest (XTHL)
takes 18
> clocks.
>
> > memory access was 1 uS
>
> The memory access time is misleading. Memory cycle time is more
useful.
> For a 2 MHz 8080, an instruction fetch cycle takes 4 cycles (2
us), and
> other memory cycles take at least 3 cycles (1.5 us).
>
> Reference: Intel 8080 Microcomputer Systems User's Manual,
September 1975,
> Intel literature part number 98-153C.





(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: A bit of history - rtstofer - Feb 11 11:49:00 2005



John

According to the time stamp on my post, it was about 11:21 PM -
seems right. Yours was about 11:30 PM - all my time (US PST).

Seems I recall calculating for some place in Australia that you were
18 hours ahead. Or, California is 6 hours ahead in time but on the
previous day. And then we have to factor in Daylight Savings Time.

I remember when I went to work in Singapore the days it took to
recover from the 16 hour time shift. It was a lot easier recovering
from the 8 hour shift when I returned.

Richard

--- In , John Kent <jekent@o...> wrote:
> Hi Richard,
>
> The 6800 I understand because of one (E) Clock = 1 cycle = 1usec.
>
> Anyway ... what are you doing up this time of night ?
> 6pm in Australia ... not sure what that is in the US,
> but I bet it's early in the morning :-)
>
> John.
>
> rtstofer wrote:
>
> >
> >The AGC has a truly bizarre architecture - clearly application
> >specific. Because the machine was so limited in memory they
> >actually wrote and interpreter and most application programs were
> >interpreted strictly to save memory. They could use lots of
lights -
> > the machine was slow enough to see them flash!
> >
> >The AGC used a 1.024 MHz clock and the fast instructions used 12
> >clock cycles. Memory cycles were 11.7 uS. The 8080 started out
at
> >2 MHz, fast instructions used 2 clock cycles, memory access was 1
uS
> >(IIRC) and the address space was much larger.
> >
> >
> >
>
> --
> http://members.optushome.com.au/jekent





(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: Re: A bit of history - Eric Smith - Feb 11 14:12:00 2005

rstofer wrote:
> I had misplaced my 8080 Microcomputer
> Systems User's Manual (Sept 1975). I have since located it - right
> on top of my "Bit-Slice Microprocessor Design" book.

By Mick & Brick? That's an excellent book. Even though people
generally don't design using 2900-series bit slice components any
more, I still recommend it to CPU designers as there is still a lot
of good information of general applicability.

Eric


Re: A bit of history


(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: A bit of history - rtstofer - Feb 11 14:20:00 2005


--- In , "Eric Smith" <eric@b...> wrote:
> rstofer wrote:
> > I had misplaced my 8080 Microcomputer
> > Systems User's Manual (Sept 1975). I have since located it -
right
> > on top of my "Bit-Slice Microprocessor Design" book.
>
> By Mick & Brick? That's an excellent book. Even though people
> generally don't design using 2900-series bit slice components any
> more, I still recommend it to CPU designers as there is still a lot
> of good information of general applicability.
>
> Eric

And I really wish I hadn't discarded the Z80 emulation app note.
They had great information on very wide control words. I have
forgotten what the assembler was but I remember having to create the
code using an 8080 macro assembler for a class I took. Great fun!



(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: Re: A bit of history - Eric Smith - Feb 11 14:31:00 2005

rtstofer wrote:
> And I really wish I hadn't discarded the Z80 emulation app note.
> They had great information on very wide control words. I have
> forgotten what the assembler was but I remember having to create the
> code using an 8080 macro assembler for a class I took. Great fun!

What Z80 emulation app note was that?

I'm familiar with two bitslice designs to emulate the 8080, one by
AMD using the 2900 series, and one by Signetics using the 3001/3002
(which they second-sourced from Intel):

http://www.bitsavers.org/pdf/amd/AMPUB-064_8080sim.pdf
http://www.bitsavers.org/pdf/signetics/Signetics8080emulMan.pdf

One of the Signetics boards sold on eBay a few years ago, but I
didn't see it until too late.

Eric


Re: A bit of history
Re: A bit of history


(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: A bit of history - rtstofer - Feb 11 14:48:00 2005


--- In , "Eric Smith" <eric@b...> wrote:
> rtstofer wrote:
> > And I really wish I hadn't discarded the Z80 emulation app note.
> > They had great information on very wide control words. I have
> > forgotten what the assembler was but I remember having to create
the
> > code using an 8080 macro assembler for a class I took. Great
fun!
>
> What Z80 emulation app note was that?

It was an AMD publication in a small paperback booklet. At one time
I had 3 or 4 of these booklets on different topics but it was so
long ago I just can't remember the details. I got far too ruthless
with my library when I moved a year or so back. > I'm familiar with two bitslice designs to emulate the 8080, one by
> AMD using the 2900 series, and one by Signetics using the 3001/3002
> (which they second-sourced from Intel):
>
> http://www.bitsavers.org/pdf/amd/AMPUB-064_8080sim.pdf
> http://www.bitsavers.org/pdf/signetics/Signetics8080emulMan.pdf
>
> One of the Signetics boards sold on eBay a few years ago, but I
> didn't see it until too late.

It is a good thing that bitsavers is out there. They have some
interesting archives. > Eric





(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: Re: A bit of history - Peter C. Wallace - Feb 11 14:52:00 2005

On Fri, 11 Feb 2005, Eric Smith wrote: > rtstofer wrote:
> > And I really wish I hadn't discarded the Z80 emulation app note.
> > They had great information on very wide control words. I have
> > forgotten what the assembler was but I remember having to create the
> > code using an 8080 macro assembler for a class I took. Great fun!
>
> What Z80 emulation app note was that?
>
> I'm familiar with two bitslice designs to emulate the 8080, one by
> AMD using the 2900 series, and one by Signetics using the 3001/3002
> (which they second-sourced from Intel):

I remember getting one of the 3002 8080 kits at some trade show a million
years ago Our programmer still has it... > http://www.bitsavers.org/pdf/amd/AMPUB-064_8080sim.pdf
> http://www.bitsavers.org/pdf/signetics/Signetics8080emulMan.pdf
>
> One of the Signetics boards sold on eBay a few years ago, but I
> didn't see it until too late.
>
> Eric >
> To post a message, send it to:
> To unsubscribe, send a blank message to:
> Yahoo! Groups Links Peter Wallace
Mesa Electronics



(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )

Re: Re: A bit of history - Eric Smith - Feb 11 16:19:00 2005

rtstofer wrote:
> It is a good thing that bitsavers is out there. They have some
> interesting archives.

Al uses my Ricoh IS520 high speed duplex scanner. :-)



(You need to be a member of fpga-cpu -- send a blank email to fpga-cpu-subscribe@yahoogroups.com )