************
Topic 12 Mon Apr 22, 1991
J.COLLINS5 [GeojoeC] (Forwarded)
Sub: 8 mhz processors for the CBM 8 bits!
I have been trying to get several Apple Hardware developers to support the
Commodore. Specifically for speed up! This is an editied exchange I had with
the online Rep for ZIP Technologies here on GEnie!
57 message(s) total.
************
------------
Category 2, Topic 12
Message 35 Mon Jan 04, 1993
C.ISAACSON1 [Dadd] at 23:35 EST
do you really think that you will ever come up with a 128 or 64 accellerator?
8s
------------
Category 2, Topic 12
Message 36 Tue Jan 05, 1993
C128.JBEE at 10:56 EST
I am 100% certain on the C-64 and almost sure on the C-128.
------------
Category 2, Topic 12
Message 37 Wed Jan 06, 1993
F.OGLE [Color 64 BBS] at 18:29 EST
Folks at the local CUGs here are screaming, Jbee - They want to know what
is up with the accelerators!
------------
Category 2, Topic 12
Message 38 Thu Jan 07, 1993
C128.JBEE at 00:14 EST
I am going ahead with them :X Once I have something concrete to post I
will, which will probably be before the end of the month. As a few of
the "insiders" know I haven't exactly been sitting still, things just
take time and I don't want to post details unless they are concrete.
Though I guarantee once I do people will be pleased :)
------------
Category 2, Topic 12
Message 39 Thu Jan 07, 1993
CBM-MARK at 22:38 EST
Does anybody know if JBEE *ever* 'sits still'? :D
We'll all be patiently (or otherwise :) waiting for you to release the
latest info when you have it!
~~Mark~~
------------
Category 2, Topic 12
Message 40 Fri Jan 15, 1993
B.ENNIS1 [Torby] at 02:47 EST
ThThe president of my local user group sent in his $50 on this accellerator. I
can't say that he has been optimistic about ever seeing anything come out of
it. Is there anything I can tell him that is of any good?
------------
Category 2, Topic 12
Message 41 Fri Jan 15, 1993
C128.JBEE at 13:49 EST
Not until the end of the month.
------------
Category 2, Topic 12
Message 42 Thu Jan 28, 1993
C128.JBEE at 19:31 EST
Bil Herd a former Commodore engineer has agreed to help me produce the
speed-up board for the C-128. Actually, Bil is the one that helped
design both the C-128 and REUs (sys 32800,123,45,6). I expect to have
him here shortly under the CBM-PRESS account.
I will also have a few other experts helping out on the project, some
which are already online ;)
Once we have everything planned out and the directions set for the
design goals I will post more here and have a RTC.
I would say at this point it is a go :)
------------
Category 2, Topic 12
Message 43 Thu Jan 28, 1993
F.OGLE [Color 64 BBS] at 19:49 EST
John, that's great news - I hope all goes well!
------------
Category 2, Topic 12
Message 44 Thu Jan 28, 1993
CBM-MARK at 22:54 EST
Hurray! Sounds like it's one step closer to completion, JBEE. Good Luck!
Hope I don't have to work when you have the RTC :(
~~Mark~~
------------
Category 2, Topic 12
Message 45 Fri Jan 29, 1993
C128.JBEE at 01:56 EST
I hope not either :)
------------
Category 2, Topic 12
Message 46 Fri Jan 29, 1993
HOWIE-CBM at 02:13 EST
This is good news!!!
Now, if we can also persuade Bil to relate a bit of Commodore lore
while he is up and about, we are all in for a huge treat. And, it is
not all that difficult to persuade Bil. :)
But I guess we should also let him have a spare moment or two so
that he can do a bit of design too.
Howie
------------
Category 2, Topic 12
Message 47 Tue Feb 02, 1993
M.LESLIE5 [Warp Sled] at 01:01 EST
That's great news! Just let me know when to get in line for it!
:)
------------
Category 2, Topic 12
Message 48 Tue Feb 02, 1993
C.OGLE2 [breadstick] at 05:56 EST
Say, JeBEE, if the outlook on the ZIP CHIP looks okay at this point, are you
still wanting people to send in deposits? Or are you going to wait and see
how the pworojusect goes first?
------------
Category 2, Topic 12
Message 49 Tue Feb 02, 1993
C128.JBEE at 21:17 EST
I don't want any more deposits or survey sheets and have been returning
the few stray deposits that have arrived (except for one from Sweden because
I know the guy had to go to a lot of trouble to send it).
I am going to delete the "zip" survey I uploaded in a few minutes.
------------
Category 2, Topic 12
Message 50 Tue Feb 02, 1993
D.WITMER2 [Snogpitch] at 23:34 EST
Looks like I've gotten into this topic a little bit late. Is there a way to
get on a waiting list, since at this point deposits are no longer being
accepted?
I've got a technical question of sorts. How does the speed up affect such
programs as telecommunication programs that involve timing to work, such as
for up/downloading, etc? Seems to me there could be a potential problem
using existing programs. Several of the local Commodore boards have been
discussing this with who knows what facts they have to back them up, and I'm
dying to find out how much software would be instantly compatible with the
zip.
------------
Category 2, Topic 12
Message 51 Wed Feb 03, 1993
C128.JBEE at 04:00 EST
There is no waiting list.
Anything that uses the 1mhz bus and CIAs for timing will be okay, software
that uses ML loops for timing will run too fast? Since my tests with
Dialogue 128 downloading at 9600 were 95%+ eff on the GEnie 9600 WATS
line I do not see where the board will speed up or improve any telecom
session worth mentioning or to justify purchasing the cartridge.
Where it will make a big improvement is in programs with a lot of processor
overhead like graphic applications (Basic 8) and things like ARC. Word
processors will run a lot faster too since the VDC can accept data far
faster than the 8502 can deliver it.
The most exciting thing (for me) will be the ability to run applications
in compiled languages such as C as fast as ML applications run now. That
will open up the possibility of a lot of new software easily ported
from other machines.
Of course things like Geos will fly :)
------------
Category 2, Topic 12
Message 52 Wed Feb 03, 1993
CBM-ED [e.g.bell] at 09:50 EST
Uh Oh!
JB> Anything that uses the 1mhz bus and CIAs for timing will be
JB> okay, software that uses ML loops for timing will run too
JB> fast? Since my tests with Dialogue 128 downloading at 9600
JB> were 95%+ eff on the GEnie 9600 WATS line I do not see
JB> where the board will speed up or improve any telecom
I know I use the jiffy timers at 160-162 for timing loops in BellTerm and
RTCMaster, including/especially for transfers. Or will that be left at the
60/second rate too? Currently that set of addresses is updated by the system
IRQ. But if the IRQ is significantly changed wrt timing, will it cause other
stuff to have problems. I am not asking like I know the answer.... I don't...
just curious.
------------
Category 2, Topic 12
Message 53 Wed Feb 03, 1993
C128.JBEE at 16:31 EST
I won't know either until we have a working prototype :)
------------
Category 2, Topic 12
Message 54 Wed Feb 03, 1993
R.KNOP1 [Rob Knop] at 21:54 EST
The normal system IRQ is triggered by the screen raster, rather than by
counting a certain number of system clock cycles. Since the screen refresh
rate can't be changed by the ZIP board (or else all of our monitors will no
longer be good enough :), I would imagine that the IRQ's will still come at 1
per 60 seconds. Which means you can get a lot more done between interrupts,
and a lot more done in an IRQ routine :).
Since the jiffy clock is updated by the IRQ routine, if my imagination that
IRQ's will still come at 1/60 seconds is correct, then software that uses the
ZP jiffy clock for timing will still probably work just fine.
By a similar token, terminal programs should still work fine. When a byte
comes in on the RS232 port, it triggers a NMI, and the actual byte receiving
is done in that NMI. Thus, the rate that byte triggers come in should not be
affected by the processor speed. And, assuming a well written terminal
program, which makes sure it isn't overfilling the buffer, there shouldn't be
any problem with sending bytes, either. (How does George Hug deal with
timing?)
Again, the terminal program will have many more clock cycles to process each
byte that comes in... visualize some pretty impessive online decompression.
Realtime GIF viewing, anybody? :D
-Rob
------------
Category 2, Topic 12
Message 55 Thu Feb 04, 1993
CBM-ED [e.g.bell] at 03:14 EST
Rob: I know the 128 IRQ is tied to the raster interrupt, but I don't think
the one in the C64 is done that way. As for normal RS232 i/o you are probably
on the money. I'm talking about file transfer protoocols where an event is
timed.... i.e. waiting for the SOH byte in all of the Xmodem derivative
protocols (Xmodem CRC/Checksum/1K/ Ymodem). If the update rate of the
addresses 160-162 are changed that is going to affect any of these protocols.
George hug uses a countdown timer with a different calibration from the
standard resident routines. That is not the whole picture, but again, my
concern was never with the garden variety i/o, which is little more than a
glorified file-handling function (my opinion) and should respond well to the
speed-up. I was just concerned about things that used the Jiffy timer. If
the interrupts are controlled at 60 per second this is probably all moot and
as you say, there will be lots more time to do things between the interrupts.
------------
Category 2, Topic 12
Message 56 Tue Feb 09, 1993
C128.JBEE at 20:37 EST
Speed-up board update:
Dennis Jarvis, the author of Basic 7.0 Internals by Abacus, an expert
on both the DOS and OS for the Commodores, who also worked on the
Turbo Master, will also be working on the new speed-up board. Dennis
will be our debugger and will be the one modifying the operating system
as needed and offering his expert advice.
A good customer who is an EE and also runs his own consulting business
will be doing the board work (soldering, chip insertion, etc).
So we have the team, we just have to rough out the plan :)
Any suggestions for the name for the new board would be appreciated.
------------
Category 2, Topic 12
Message 57 Wed Feb 10, 1993
HOWIE-CBM at 05:48 EST
