Not sure if this is accessible by everyone, but IEEE spectrum has a great article entitled "25 Microchips That Shook the World". It's well worth a read if you are even only a little bit geeky. 
25 chips "that shook the world"
- Thread starter Simon F
- Start date
Apart from an attempt with the RCA 1802, the 6502 was my first experience with programming in machine code. I say machine code because on the Apple ][ there was a built-in disassembler but not an assembler.
Simon
Jawed
Legend
Yep, my first machine code too. I had neither assembler nor disassembler when I started, doing everything with pokes and a call and my trusty copy of Rodnay Zak's Programming the 6502. I rewrote a 3D maze navigation game to draw the walls in machine code rather than Basic. The performance increase made my jaw drop. Of course I used the extra performance to increase the view distance.
This was on an Ohio Superboard II, with 24x24 character graphics. Walls drawn with _ | \ and / "characters" in the extended ASCII, the characters being 8x8 dots, unlike many machines of that era with rectangular character matrices.
Jawed
This was on an Ohio Superboard II, with 24x24 character graphics. Walls drawn with _ | \ and / "characters" in the extended ASCII, the characters being 8x8 dots, unlike many machines of that era with rectangular character matrices.
Jawed
Two chips in there are highly relevant to me.
I got the bug on the 6502 (Microtan 65 anyone ?)
first project was typing in, via a 16 key keypad, a HEX dump of a space invaders program from computing today (64x64 screen resolution). And this was BEFORE I had the tape backup option, every time you wanted to play it, you had to type in 1.4K or whatever of hex code.
I made (make) my living on 8080/Z80 archeitecture and its derivatives.
<---looks up to the bookshelf to see it still sitting there
Ummm.. on that Microtan 65, when I got the "extended 10K basic" for it, if you did a key combination it came up with "written by Weiland and Gates".
I got the bug on the 6502 (Microtan 65 anyone ?)
first project was typing in, via a 16 key keypad, a HEX dump of a space invaders program from computing today (64x64 screen resolution). And this was BEFORE I had the tape backup option, every time you wanted to play it, you had to type in 1.4K or whatever of hex code.
I made (make) my living on 8080/Z80 archeitecture and its derivatives.
JAWED said:
<---looks up to the bookshelf to see it still sitting there
Ummm.. on that Microtan 65, when I got the "extended 10K basic" for it, if you did a key combination it came up with "written by Weiland and Gates".
Apart from an attempt with the RCA 1802, the 6502 was my first experience with programming in machine code. I say machine code because on the Apple ][ there was a built-in disassembler but not an assembler.
I started out on a C64 with a 6510 which is just a 6502 with a few IOs integrated. Initially programming was done in machine code with a runtime monitor using mnemonics. Later used Commodores awful slow-as-molasses two pass floppy disc based assembler until finally using Turbo Assembler. Lots of demos and small games.
Cheers
Whoah, yields must have been REALLY bad back in those days...?
But yea, I started on a 6502 aswell, with the C64. I had a ZX81 before that (Z80 CPU), but I never used anything but BASIC on that one.
After that I went 68000 on the Amiga. I remember the frustration when I first did 8086 code on a PC. I felt like it was more like the 6502 than the 68000, even though the PC was far more modern (not to mention expensive).
Great bit of nostalgia, that article
That or it was SIGNIFICANTLY smaller?
Silent_Buddha
Legend
I'm not sure the 6502 was that significantly smaller than the 8080. And 25 USD versus 200 USD is quite a gap. Although granted that could have just been Intel and Motorola (with the 6800) milking incredibly large margins.
In today's money that would be like the 6502 at 100 USD versus the Intel 8080 at almost 800 USD.
Quite nicely showing why it was such a huge hit.
Regards,
SB
In today's money that would be like the 6502 at 100 USD versus the Intel 8080 at almost 800 USD.
Quite nicely showing why it was such a huge hit.
Regards,
SB
According to Wikipedia, the 8080 had about 6000 transistors, and a die size of about 20 mm^2.
The 6502 had about 4000 transistors. It doesn't say what the die size was, but I doubt that the size alone would be able to account for 10 times better yields.
I interpreted the statement as the fabrication process having some revolutionary new technology that made it far more reliable and consistent.
That would also explain why MOS was able to sell it at much lower prices than Intel.
The 6502 had about 4000 transistors. It doesn't say what the die size was, but I doubt that the size alone would be able to account for 10 times better yields.
I interpreted the statement as the fabrication process having some revolutionary new technology that made it far more reliable and consistent.
That would also explain why MOS was able to sell it at much lower prices than Intel.
From my memory, based on what I red in the On the Edge book:
Back in the day the masks for lithography (I hope thats the right word) were cut out from a large sheet of plastic (5m² or so) by hand and then progressively scaled down. Needless to say, cutting out 3-5 layers of plastic with thousands of small elemenents was a painful job and errorprone job
MOS was the first who used a lithography procedure that worked without the masks having physical contact to the wafer. All others had their masks wear down, making yields worse and needed to regulary create new masks.
Instead MOS only needed to create 1 mask and could spend more time making this mask the best they could.
Back in the day the masks for lithography (I hope thats the right word) were cut out from a large sheet of plastic (5m² or so) by hand and then progressively scaled down. Needless to say, cutting out 3-5 layers of plastic with thousands of small elemenents was a painful job and errorprone job
MOS was the first who used a lithography procedure that worked without the masks having physical contact to the wafer. All others had their masks wear down, making yields worse and needed to regulary create new masks.
Instead MOS only needed to create 1 mask and could spend more time making this mask the best they could.
What no GPUs?
My guess is that history will look on GPUs like it does the 8087.
Meaning as a co-processor? I dunno, I tend to put GPUs in the same category as DSPs, which were on the list.
Did you mean "what has happened to it?" or "why isn't it listed?". If the latter, I'd probably former answers that.
Did you mean "what has happened to it?" or "why isn't it listed?". If the latter, I'd probably former answers that.
I did mean the former and, frankly, can't think of anything in that genre that has made a dent; even on headlines. But as I said I know very little about it. Hoping someone can correct me!
