I think that the ARM as originally conceived by Acorn Computers was easy to programme in assembler - the chief designer was a fan of the 6502 and wanted something that was as easy to use. I bought a couple of the original chips - I think they were in the PLCC package - but never got round to using them.
Turned out that a BASIC interpreter does not need a proper, big, stack it builds it's own for the program being interpreted. Boom, there is the PET, Vic 20 and C64.
Acorn, really had no choice but to make a RISC machine. They only had a few chip designers, perhaps only one, Sophie Wilson. So it had to be small and simple and doable in the time/budget available.
Distributed Intelligence, was the term Chuck Peddle used. Mostly in the context of solving point of sale problems, which were significant and growing in the 70's.
On the 6502, the Zero Page was intended for things like extra registers, stacks and such.
And it's still being used in that basic way today. Western Design Center actually released a new 6502 family development board. Spiffy! I sort of want one.
But, I have 6502 computers now. So, just use those.
On a side note, "The Amp Hour" is a great podcast. Well produced, informative and entertaining. Recommended.
Interesting... I don't know much about it. Was a great interview though. He's got some interesting perspective, and is still active. Love that.
But, a Propeller out of 6502's...
Say the 6502 remains a 6502, load store, etc...
A COG would be 2K of RAM, addressed $0-$1ff. So that's nice. The Stack and Zero Page would be in the map and useful. One could make a transfer register for byte, word, long moves to and from HUB. All of the little indexing modes in the 6502 work great for the COG address space too. No worries about COG code. And there is the page bug, not likely to come up much. Clean.
If it were done how the Prop is done, then firing off a COG would involve a HUB to COG copy, just as we have now, then some launch into the program, just like we have now. The thing that makes a Prop is that COG isolation, and the HUB being shared.
COG addressing would have to be byte, which is kind of nice. Unless there is some hardware, "move it for me, given an address" function, it's all bytes to and from the HUB, which is slow. I would add that capability.
Wonder just how fast a 6502 type core could run?
Back in the day, I saw a dual 6502 chip in one of the databooks, Rockwell I believe. It was two 6502's that shared the same package, and they were interleaved, running every other clock cycle. Not sure that one was ever produced. Never saw a mention of it, outside that databook. Thought it might be fun to modify my Apple to use it at the time. Dual core Apple ][
We have the P1 code. Maybe somebody someday will attempt something crazy like a 6502 / 6809 / Z80 Prop type thing. Would be a fun curio.
No idea, but if a bunch of university students can get a 32 bit a RISC V core made that runs at giga hertz plus speeds now a days then I guess a 6502 core can do the same.
Give each core 32K or so of RAM, like the Adepteva Epiphany chip's 16 cores, and you have a substantial real-time engine.
Just about any CPU could be run at 2 GHz if you are willing to spend the money on a mask set (read that's big money). The real issue is memory access, in a 6502 case, you could just stick it all on chip and run about the same speed. But what can you do in 64K anymore? Web browser - no, Python - no, Linux - no** ...
On the FPGA side, around 10 years ago we found some HDL for a small PIC that we compiled a few copies into a XIlinx part and could easily run them at 200 MHz. They made great programmable state machines.
** no means I am getting older and am no longer willing to wait hours for a prompt or a page refresh.
Here in New England, we have put Spring off for another two weeks, (unofficially)
Snowed Saturday and Sunday, and is going to snow tonight. Daffodils are trying to push through.
The Daffodils, Bluebonnets and all kinds of things are in full bloom here. However, I'm more concerned about the end of Spring, and 6 months later, the end of Fall in Rocklin, CA.
200Mhz! That's going to center in on about 50 MIPS average. Insane!
And the original 2MHz (iirc) was used to produce video in text and low resolution graphics terminals. Imagine what some creative person could do with 200MHz.
I think that the ARM as originally conceived by Acorn Computers was easy to programme in assembler - the chief designer was a fan of the 6502 and wanted something that was as easy to use.
Rightly or wrongly (apparently wrongly), I've long blamed Sophie Wilson for turning a simple concept into something complicated. Perhaps it was Marketing instead. They are invariably a meddlesome bunch!
I have read a couple of times that Wilson and or others at Acorn were looking around for CPU for the next gen computers they would build. Being familiar with simplicity of the 6502 they were not impressed with the Intel, Motorola and other offerings available at the time. One of them happened to visit a company in the USA and there they saw how it was possible for a small team to be designing chips. "We can do that they thought".
And so the ARM was born, inspired by the principles of the 6502 with a sensible 32 bit data width.
The resulting Archimedes computer was for a while by far the fastest personal computer one could get. How I lusted after it.
That's right. Wilson and his/her boss visited Western Design Center near Phoenix and saw that pretty much one guy was maintaining the 6502 IP with an HDL. No huge capitalization needed.
I have read a couple of times that Wilson and or others at Acorn were looking around for CPU for the next gen computers they would build. Being familiar with simplicity of the 6502 they were not impressed with the Intel, Motorola and other offerings available at the time. One of them happened to visit a company in the USA and there they saw how it was possible for a small team to be designing chips. "We can do that they thought".
And so the ARM was born, inspired by the principles of the 6502 with a sensible 32 bit data width.
The resulting Archimedes computer was for a while by far the fastest personal computer one could get. How I lusted after it.
I still liked the Amiga better with it's custom grahics controller .... graphics today is still clunky compared to the good old Amiga days ..
... was for a while by far the fastest personal computer one could get. How I lusted after it.
Funny how terminology has changed, and only very recently too. Personal Computer, generally capitalised, was always reserved for the clone market and related Wintel monoculture. Only in the last few years have I seen it used in reference beyond that.
Previously, microcomputers outside of the PC market were just called home or pocket computers with the business oriented ones called workstations or handhelds.
I still liked the Amiga better with it's custom grahics controller .... graphics today is still clunky compared to the good old Amiga days ..
I'm amazed the demo scene still exists. The cracker intros were a joy to behold, and left me in awe many times ... spent more time watching them than playing the games they came with. Lol, spent even more time just going to the copy fests.
It does, and they are trying really hard to get some recurring parties and fans going in the US.
Frankly, I love the demo scene, and have followed since the first intros I got on disk years ago. The bigger and better parties are streamed and very well produced. I'll put on on for the weekend, paying attention to some productions and just letting others happen.
A Propeller has been seen in "Wild" a coupla times. There is a scene board being marketed to people now. (should have been a Prop, but isn't)
Amiga gets broken down into AGA and Pre-AGA (golden age, or should I say Copper) and sees several productions per year. C64 actually sees 5-10 sometimes!
Our own Baggers submitted Speccy demo stuff to a recent party.
The whole thing is doing OK, but it needs to pick up some more younger people. Actually, scratch that. The scene just needs more people.
Funny how terminology has changed, and only very recently too. Personal Computer, generally capitalised, was always reserved for the clone market and related Wintel monoculture. Only in the last few years have I seen it used in reference beyond that.
I did not capitalize "personal computer" on purpose. The term was widely used way before the IBM PC and was familiar to us youngsters although I don't recall using it much.
We had other names for the IBM PC when it arrived, most of which are not fit to be repeated here:)
I'd only ever heard of mini/micro-computers in general type references dating back then. Even when I went to work industrially, colleagues used same terms of mini/micro to separate out the counter top equipment from the central data-processing machines. And of course PC definitely did mean compatibles and clones by that stage, being mid-1980's. It was almost a specific brand.
"personal computer" was definitely in use before the IBM 5150, that is why so many (including me) reacted negatively to IBM taking it for itself. The naming was discussed in magazines too. Now there are scans of many of the old mags (although not for the great PCW which I subscribed to from 1977) and it's possible to go and check up on this. And yes, I've seen "personal computer" recently when flicking through scans. For SBC type boards too IIRC.
Comments
Turned out that a BASIC interpreter does not need a proper, big, stack it builds it's own for the program being interpreted. Boom, there is the PET, Vic 20 and C64.
Acorn, really had no choice but to make a RISC machine. They only had a few chip designers, perhaps only one, Sophie Wilson. So it had to be small and simple and doable in the time/budget available.
On the 6502, the Zero Page was intended for things like extra registers, stacks and such.
And it's still being used in that basic way today. Western Design Center actually released a new 6502 family development board. Spiffy! I sort of want one.
But, I have 6502 computers now. So, just use those.
On a side note, "The Amp Hour" is a great podcast. Well produced, informative and entertaining. Recommended.
But, a Propeller out of 6502's...
Say the 6502 remains a 6502, load store, etc...
A COG would be 2K of RAM, addressed $0-$1ff. So that's nice. The Stack and Zero Page would be in the map and useful. One could make a transfer register for byte, word, long moves to and from HUB. All of the little indexing modes in the 6502 work great for the COG address space too. No worries about COG code. And there is the page bug, not likely to come up much. Clean.
If it were done how the Prop is done, then firing off a COG would involve a HUB to COG copy, just as we have now, then some launch into the program, just like we have now. The thing that makes a Prop is that COG isolation, and the HUB being shared.
COG addressing would have to be byte, which is kind of nice. Unless there is some hardware, "move it for me, given an address" function, it's all bytes to and from the HUB, which is slow. I would add that capability.
Wonder just how fast a 6502 type core could run?
Back in the day, I saw a dual 6502 chip in one of the databooks, Rockwell I believe. It was two 6502's that shared the same package, and they were interleaved, running every other clock cycle. Not sure that one was ever produced. Never saw a mention of it, outside that databook. Thought it might be fun to modify my Apple to use it at the time. Dual core Apple ][
We have the P1 code. Maybe somebody someday will attempt something crazy like a 6502 / 6809 / Z80 Prop type thing. Would be a fun curio.
Give each core 32K or so of RAM, like the Adepteva Epiphany chip's 16 cores, and you have a substantial real-time engine.
On the FPGA side, around 10 years ago we found some HDL for a small PIC that we compiled a few copies into a XIlinx part and could easily run them at 200 MHz. They made great programmable state machines.
** no means I am getting older and am no longer willing to wait hours for a prompt or a page refresh.
Here in New England, we have put Spring off for another two weeks, (unofficially)
Snowed Saturday and Sunday, and is going to snow tonight. Daffodils are trying to push through.
Of course not. It's a P1 6502 style.
Yeah, Chip said the same thing. Getting the P1 at a Ghz is just a hand full of zeroes and a little time away.
Yeah, that would be Spring of 2015, this time.
Edit: Reference: http://www.westerndesigncenter.com/wdc/
And the original 2MHz (iirc) was used to produce video in text and low resolution graphics terminals. Imagine what some creative person could do with 200MHz.
Rightly or wrongly (apparently wrongly), I've long blamed Sophie Wilson for turning a simple concept into something complicated. Perhaps it was Marketing instead. They are invariably a meddlesome bunch!
And so the ARM was born, inspired by the principles of the 6502 with a sensible 32 bit data width.
The resulting Archimedes computer was for a while by far the fastest personal computer one could get. How I lusted after it.
Boggles the mind to think where that has led.
I still liked the Amiga better with it's custom grahics controller .... graphics today is still clunky compared to the good old Amiga days ..
Funny how terminology has changed, and only very recently too. Personal Computer, generally capitalised, was always reserved for the clone market and related Wintel monoculture. Only in the last few years have I seen it used in reference beyond that.
Previously, microcomputers outside of the PC market were just called home or pocket computers with the business oriented ones called workstations or handhelds.
Microcomputer is still a valid label.
... and another six months!
I'm amazed the demo scene still exists. The cracker intros were a joy to behold, and left me in awe many times ... spent more time watching them than playing the games they came with. Lol, spent even more time just going to the copy fests.
Frankly, I love the demo scene, and have followed since the first intros I got on disk years ago. The bigger and better parties are streamed and very well produced. I'll put on on for the weekend, paying attention to some productions and just letting others happen.
A Propeller has been seen in "Wild" a coupla times. There is a scene board being marketed to people now. (should have been a Prop, but isn't)
Amiga gets broken down into AGA and Pre-AGA (golden age, or should I say Copper) and sees several productions per year. C64 actually sees 5-10 sometimes!
Our own Baggers submitted Speccy demo stuff to a recent party.
The whole thing is doing OK, but it needs to pick up some more younger people. Actually, scratch that. The scene just needs more people.
We had other names for the IBM PC when it arrived, most of which are not fit to be repeated here:)
Checking on it now I see that "personal computer" dates back to a 1962 New York Times article.
http://en.wikipedia.org/wiki/History_of_personal_computers#Etymology