A Matter of RAM

8 posts / 0 new
Last post
gsmcten's picture
Offline
Last seen: 5 years 8 months ago
Joined: Oct 4 2005 - 18:52
Posts: 2629
A Matter of RAM

I am not sure if this subject has been touched upon or not, but as I was driving back home from the store this morning (pondering the universe as we know it), I wondered why there is a limit on the amount of RAM accessable on an Apple II?

I know that somewhere I have read that the IIgs can only access 8 Meg of RAM.
Most RAM cards are 256 K to 1 Meg, with the exception of 1 AE Card that has 1 Meg Main RAM with a daughterboard of another 1 Meg.

Apple II users have made great strides the last 15 years with Mass Storage, Speed, and many other areas, except the RAM Limit.

I would be interested to hear the thoughts of everyone about this subject.

Is there a way to access RAM greater than 1, or 2 Meg on an Apple II, II Plus, or IIe?
Is there a way to access RAM Greater than 8 Meg on a IIgs?

Is there anyone trying to develope a RAM Card like this, with say maybe 1 Gig?

Wouldn't a 1 or 2 Gig Accessable RAM Card for the Apple II Series be nice?

Steven Smile

mmphosis's picture
Offline
Last seen: 2 hours 32 min ago
Joined: Aug 18 2005 - 16:26
Posts: 432
Re: A Matter of RAM

The 65816 has a 24−bit Address Bus and thus should be able to access 16Mbyte Memory and I/O Space. I remember trying to figure out the cost of putting 8Meg of RAM in the IIGS and it was astronomical at the time.

I am pretty sure there is a way to build 16-byte or 128-bit bank switch addressing on an Apple II card. I know more about software, but I think that there are various lines that allow cards to do things with the addressing of RAM and ROM in the Apple II.

Bank switching cards I've seen in the past have been limited to bank switching using only a single byte in the I/O address space. I propose the following:

Bank switching using card I/O via C0x0-C0xF -- 16-bytes for a 128-bit bank switch!

  • Access the RAM through the Cx00-CxFF range for 256 bytes at a time
  • Access the RAM through the C800-CFFE ranges for 2K bytes at a time.
  • Access the RAM through the D000-FFFF range by disabling the ROM and be able to access 12K at a time.
  • And, finally be able to address 64K of RAM at a time.

With fewer limits on RAM, I know that very long reams of code could be generated that would run very fast. It would only need to bank switch every so often to swap in more RAM/code.

Here is an example of generated code that runs fast, but consumes lots of memory space:

NEW
0 HIMEM: 5608
1DATA5532563300922139021390213932135045238263536503037252627503845615305817312092293902932835845478875645836575037148845550351083966132653706165276445377489621322135821322334213502130051520282036
2DATAQLNZQLNZQAQQDSAQRDSAQQDSAQVDSAQXCKDNAPFXANAQXXANANXNXNXQXNXCQXKXNZQXKXNZQCQODMAQODMAUUYQXCKATAUVQXCKMNXQXKANXUAUTCQXKENXUMUSQJHSAHFQZTXRDFQZTAOCQZTAOBQHDSAPDSAQADSANZNZKDSAQZDSAXZQZOZXZUAXZJ
3 READ L$: READ H$
4 FOR I = 1 TO LEN (L$)
5POKE767+I,10*(ASC(MID$(H$,I,1))-65)+VAL(MID$(L$,I,1))
6 NEXT
RUN

HGR : CALL 768: CALL 5608

speedyG's picture
Offline
Last seen: 4 years 9 months ago
Joined: Nov 16 2011 - 07:45
Posts: 2493
Re: A Matter of RAM

Hello gsmcten,
in fact the only rule is that you can do everything you like - you just must be able to pay for it....

one of the most common problems with this topic is the so called TP-area...
thats the area remaining left after the bankswitching program is loaded and avoiding the use of the so called reserved areas
like the graphic-display areas and HGR pages as well as the slot areas....
but its not only bankswitching...
you also have to pass over that info of availiable memory to the program so that istmight use it.....
nearly 98 % of all programs written for the II series would never make use of that area and those programs,
that could make use of it, like Appleworks need special teatment to recognize the availiable memory....

AE engineers paid a very high amount to the guys that developed Appleworks to recieve the internal code
of Appleworks to find out how it determines availiable memory-space....
so here´s the true graveyard for the memoryexpanders....

and programming the expanders for use with the operation-systems ( are you thinking about some kind of
"better" OS than Kontiki to load modern websites and modern browser ??? ) is a very timeeating task....
you can spend several years with the development of a new OS to use such amount of memory.....
and if you don´t want it to be reduced to "snail-racing" you would have to do most of that in assembler and integrate routines of the ROM
- otherwise you might start cooking a meal and finally after you have eaten it
- your program probably might have loaded successfully....

sincerly speedyG

gsmcten's picture
Offline
Last seen: 5 years 8 months ago
Joined: Oct 4 2005 - 18:52
Posts: 2629
Re: A Matter of RAM

Speedy and All,

LOL
No. I'm not a developer in any shape or form.
It just seemed to me when I wrote the original Topic, that with all the advancements that have been made with speed and storage for the II's, that someone would have come up with a way to break the Memory Barrier.
(Kinda like breaking the sound barrier, only better) Smile

Thats why I tossed it out here. It may get someone to thinking about how it could be done.

Steven Smile

Dog Cow's picture
Offline
Last seen: 4 years 9 months ago
Joined: Dec 11 2008 - 16:26
Posts: 554
Re: A Matter of RAM

I am not sure if this subject has been touched upon or not, but as I was driving back home from the store this morning (pondering the universe as we know it), I wondered why there is a limit on the amount of RAM accessable on an Apple II?

All computers have a limitation on the amount of memory which they may address, even computers today, though those limits are quite high.

The 6502 CPU in the early, 8-bit Apple II computers had only a 16-bit address line, meaning that a memory address was only 2 bytes. Two bytes allows for only 65,536 memory locations.

As you probably know, in the standard Apple II and II Plus, 12K of the address space is for the ROM, 512 bytes are reserved for page zero and the stack, and the 4K at $C000 is reserved for I/O access. That gives the programmer access to roughly 48K of RAM.

gsmcten's picture
Offline
Last seen: 5 years 8 months ago
Joined: Oct 4 2005 - 18:52
Posts: 2629
Re: A Matter of RAM

Dogcow,

Yep. Understand all of that.

I tossed this out to the critters to get a conversation to brew.

It's kind of like:

"Don't tell me the limitations. Tell me the possibilities."

Is it possible?
Is there a workaround (either hardware or software)?

Men looked at the Moon once. Then they went there, right?

Steven Smile

speedyG's picture
Offline
Last seen: 4 years 9 months ago
Joined: Nov 16 2011 - 07:45
Posts: 2493
Re: A Matter of RAM

Hello Steven,
just to remain in your terms...:
we all looked to the Limits in 1979 to 1983...
1984 -> the guys from Applied Engineering stepped first time on the moon with the RAMWORKS
then they returned to the moon with the RAMWORKS II
after that they arrived again in the (permanent?) moonbase with RAMWORKS III
and after the IIGS arrived in the market ( comparing that to a look at the mars )
the guys from AE again had been the first to put their feet on the ground of the Mars with a own RAM-Expansion-Card
the problem was same as to the NASA with the Space program ..... you need somebody to pay for the trip......
thats why i claimed initially:
the problems are known,
the solutions are existing....
but if you want more.... you got to pay for it..... and the price may probably not have any relation to the advance....
also the NASA spent several millions to rockets exploding in the sky and arriving in nowhere..... thats price of development.....
i just wonder that people/customers allways only think about of the price of a item - but not realizing the price to pay to get there....
sincerely
speedyG

gsmcten's picture
Offline
Last seen: 5 years 8 months ago
Joined: Oct 4 2005 - 18:52
Posts: 2629
Re: A Matter of RAM

Speedy and All,

Yes, but is the cost prohibitive to a project like this?

Log in or register to post comments