6 Jan
2010
6 Jan
'10
7:28 a.m.
I noticed that there's an article in the december circuit cellar that
talks about emulating an Apple II+ in an FPGA. It was an interesting
article, although I only understood about 20% of it. As a software
guy though, the hardware "code" has me interested. Maybe this would
be an easier and less time-consuming way for me to learn about
hardware.
brian
6 Jan
6 Jan
8:27 a.m.
Brian Lanning wrote:
I noticed that there's an article in the december
circuit cellar that
talks about emulating an Apple II+ in an FPGA. It was an interesting
article, although I only understood about 20% of it. As a software
guy though, the hardware "code" has me interested. Maybe this would
be an easier and less time-consuming way for me to learn about
hardware.
brian
Brian,
I've got to subscribe -- sounds like a neat article. I recently
subscribed to Nuts and Volts, and I like the fact that you get access to
all of the back issues back to like 2004 for free included w/ your
subscription.
The whole retro-computing thing in an FPGA floats my boat. Very neat
stuff. Lots of people have been reverse-engineering or duplicating the
functionality of these old custom chips which is fantastic. Plus I love
how it scales. Got two custom chips? 10 custom chips? The logic will
fit into a small FPGA. And then throw the soft-core processor onto the
same FPGA and you have the system-on-a-chip idea. And there's little
wiring to do because everything is internal.
As far as less time-consuming goes, I wouldn't count on it. I've had an
FPGA board for awhile and there is a learning curve there. I've enjoyed
the time I've spent on it, but like anything else, there is a fair bit
of complexity if you hope to understand how everything works. The
installed Xilinx(in my case) ISE software is huge, and requires a decent
machine to give reasonable performance. The fact that many tools are
separate programs, while (mostly) launched from the same place, leads to
minor integration problems. Like tools that are command line only, and
need 8.3 filenames. Or other tools that can't handle spaces in
filenames (like "Documents and Settings", ie) And then there's
documentation, which is mostly good, but BIIIIG. Like 600 pages for a
memory controller.
I'm not trying to sway your decision. Playing w/ this hardware is lots
of fun --- just want to set your expectations accordingly.
Keith
8:52 a.m.
On Wed, Jan 6, 2010 at 10:27 AM, Keith Monahan <keithvz at verizon.net> wrote:
I've got to subscribe -- sounds like a neat
article. ?I recently subscribed
to Nuts and Volts, and I like the fact that you get access to all of the
back issues back to like 2004 for free included w/ your subscription.
I was an EE major my first year of college before switching to CS. It
was probably a good career move, but I miss playing with electronics.
I really don't need another hobby though.
The whole retro-computing thing in an FPGA floats my
boat. ?Very neat stuff.
?Lots of people have been reverse-engineering or duplicating the
functionality of these old custom chips which is fantastic. ?Plus I love how
it scales. ?Got two custom chips? 10 custom chips? ?The logic will fit into
a small FPGA. ?And then throw the soft-core processor onto the same FPGA and
you have the system-on-a-chip idea. ?And there's little wiring to do because
everything is internal.
That's how the minimig is. Someone reimplemented the amiga custom
chips and a 68000 in an FPGA. I believe the new c64 chameleon works
like this also.
As far as less time-consuming goes, I wouldn't
count on it.
I agree. There's so much to learn and nowhere near enough time. I'd
like to learn to weld also, but that's not likely to happen any time
soon either.
I'll probably stick with tinkering with 6502 machine language. At
least I'm already equipped to do that. The learning curve should be
easy.
brian
9:40 a.m.
Brian Lanning wrote:
That's how the minimig is. Someone reimplemented
the amiga custom
chips and a 68000 in an FPGA. I believe the new c64 chameleon works
like this also.
I am a proud minimig owner. :)
The minimig is awesome. It was actually the inspiration for me to learn
verilog and FPGAs.
I agree. There's so much to learn and nowhere
near enough time. I'd
like to learn to weld also, but that's not likely to happen any time
soon either.
It's hard to really describe in words how DEEP and vast the FPGA world
is. You'd think that learning your HDL is sort of 90% of the work. But
it's really not true, learning Verilog is the easy part. :) The hard
part is learning how all the tools work together to yield the desired
result.
I'll probably stick with tinkering with 6502
machine language. At
least I'm already equipped to do that. The learning curve should be
easy.
brian
I've found that stuff straight-forward as well. Designing for FPGAs
definitely takes a different mindset. You don't really write code, you
design hardware. Practically everything is concurrent and simultaneous
--- so the concept of one instruction after the other simply doesn't
exist (for the most part). It's weird to see a block of code that,
ummm, is always running. I mean, always there. It's good and it's bad.
And your registers are always the exact size you specify --- doesn't
matter if it is 3-bits wide or 100 bits wide. And there's no
limitations except physical ones.....
My current project is going to use picoblaze (8-bit microcontroller core
) with hardware for a timer and time-critical things.
Keith
1:29 p.m.
My current project is going to use picoblaze (8-bit
microcontroller core
) with hardware for a timer and time-critical things.
Hm, that's a quite
different thing than building an Apple computer in an FPGA :-)
--
http://www.hachti.de
10:46 a.m.
While I'm on my podium, one more quick thing:
(I know this isn't your word perse, so I'm not picking on you Brian)
but I think the word "emulate" is really walking a thin line when we
talk about this stuff. The minimig, for instance, is really an Amiga
clone. It doesn't "emulate" a processor -- it either has a physical
hardware processor or a processor core in the FPGA. The hardware
actually exists physically. The custom chips logic is physically
instantiated and temporarily created within the FPGA.
All this stuff is sort of "better than", or one step above, emulation.
Here's the best part: let's say you don't like how the original amiga
bus worked. You can take the existing design, and reimplement an
entirely new bus, reconnect the various logic to the new bus, and off
you go. What's amazing is that you are simply modifying code to make
all this happen. Never touched a PCB design tool, wire wrap tool, wires
on a breadboard, etc.
Best yet, say you don't even have (any) hardware in front of you -- you
can simulate your newly designed creation and look at logic waveforms to
test your creation. Does it do what you intended? Is the timing right?
Amazing stuff here.
Keith
10:54 a.m.
On Wed, Jan 6, 2010 at 12:46 PM, Keith Monahan <keithvz at verizon.net> wrote:
but I think the word "emulate" is really
walking a thin line when we talk
about this stuff. ?The minimig, for instance, is really an Amiga clone. ?It
doesn't "emulate" a processor -- it either has a physical hardware
processor
or a processor core in the FPGA. ?The hardware actually exists physically.
?The custom chips logic is physically instantiated and temporarily created
within the FPGA.
I have software developer brain damage. :-)
brian
11:57 a.m.
* On Wed, Jan 06, 2010 at 12:54:57PM -0600, Brian Lanning <brianlanning at
gmail.com> wrote:
On Wed, Jan 6, 2010 at 12:46 PM, Keith Monahan
<keithvz at verizon.net> wrote:
I like that phrase :-)
This was really the biggest challenge to me when I started playing with
FPGAs via an inexpensive Digilent board. Before 2007, I had never really
touched hardware. I knew what a resistor was, that was about it. But I've
been programming since I first laid hands on a Commodore 64, so those
concepts were deeply ingrained. It took a little while - and a bunch of
synthesis errors - before I really understood that VHDL and Verilog are not
programming languages. They're hardware description languages. I had to get
that software developer brain damage out of my head and start anew!
The good news is that the FPGA was a gateway drug. When I had my epiphany
about hardware description langauges, I realized I'd better understand a
little something about hardware, and that led to my buying a breadboard and
some components. Now I have a fully stocked electronics workbench that's
led to a lot of fun tinkering, and a much richer understanding of the
computers that I've been collecting all these years.
but I think the word "emulate" is
really walking a thin line when we talk
about this stuff. ?The minimig, for instance, is really an Amiga clone. ?It
doesn't "emulate" a processor -- it either has a physical hardware
processor
or a processor core in the FPGA. ?The hardware actually exists physically.
?The custom chips logic is physically instantiated and temporarily created
within the FPGA.
I have software developer brain damage. :-) brian
-Seth
9 Jan
9 Jan
11:59 a.m.
On Wed, Jan 6, 2010 at 4:28 PM, Brian Lanning <brianlanning at gmail.com>wrote:
I noticed that there's an article in the december
circuit cellar that
talks about emulating an Apple II+ in an FPGA. It was an interesting
article, although I only understood about 20% of it. As a software
guy though, the hardware "code" has me interested. Maybe this would
be an easier and less time-consuming way for me to learn about
hardware.
Go for it:
http://www1.cs.columbia.edu/~sedwards/apple2fpga/
http://alexfreed.com/FPGApple/
http://alexfreed.com/FPGApple/revisited.html
http://www.applelogic.org/
--
Regards,
Torfinn Ingolfsen
5243
days inactive
5246
days old
8 comments
5 participants
participants (5)
-
brianlanning＠gmail.com -
hachti＠hachti.de -
keithvz＠verizon.net -
sethm＠loomcom.com -
tingox＠gmail.com