> "Core" memory was invented in the late 40s, but didn't come into wide use until
> the early 50s.  I think the IBM 360s used core memory.

How about drum memory? Or the various delay lines?

> The IBM 370s and all later machines used integrated circuits for memory.

As in, whole memory chunks on a single IC? Or just two latches per IC? 
Or...?

> We added 1MB of memory to an IBM 370/168 in the mid 70s.  Price? $50000.

Ouch! o_O

It still somewhat blows my mind that you could do anything useful with 
so little memory. Presumably for processing large datasets, most of the 
data at any one time would be in secondary storage?

> There were several advantages to core memory.

Yes, so I see.

> When I was at university in the mid to late 60s, I took a FORTRAN programming
> course.  Punch program into cards, place in card tray, come back 4 to 6 hours
> later and pick up your card deck and output listing.  Fix your typos and bugs
> and repeat the above process until the program works.

Heh, sounds line fun. According to the account I just read, you actually 
put one statement per punch card. Is that right? I always assumed that 
each card just held X characters of data, and you type until the card is 
full, then move to the next one. I didn't realise there was an actual 
"significance" to card bounderies. Or is that just for the benefit of 
the humans?

> Where I worked cards were used until the mid 1980s.

Wow. I had no idea it persisted that long. In the mid 1980s, I had a C64 
to play with, which stored data on audio tape - for more data than any 
stack of punch cards. And I don't think the C64 was even particularly 
expensive... (I know of a few people who actually used the C64 for work 
purposes, would you believe.)

> The fun began after you punched in your 1000 card program and dropped
> the deck on the way to the card reader.  After you did that a couple of times,
> you started adding sequence numbers to the deck.

Hell yeah! ;-)

>> Was there ever a "punched tape" medium similar to punch cards?
>
> Yes. Several different varieties of tape.

OK. I guess each manufacturer probably had their own style.

> I once had to write a driver for the IBM 1056 paper tape reader on
> an IBM 370/158 because IBM did not support that device on that machine.  That
> was kind of fun to do. That was in the early 70s?

Hmm, OK.

What sort of access speeds do you get for reading or writing punched 
card or tape?

>> Similarly, you hear people talk about the VAX, the PDP, the varouis IBM
>> mainframes and Cray supercomputers. Does anybody know the timeline for
>> these, the technologies used and the basic design and performance details?
> 
> I think you will have to look it up.  There is just too much information to
> enter here.  I have work exclusively on IBM machines my entire career.  So I
> can't say anything about other manufacturers.

Fair enough.

> Very briefly IBM 360 series, transistor circuits, core memory,
> mid 60s to early 70s.

As in, discrete transistors on PCBs?

Any ideas on typical memory capacity / clock speed? Data path widths?

> Early IBM 370 series in
> the early 70s, integrated circuits, to later 370 series with higher density
> circuits in the mid 80s.

Are we talking ICs with single logic gates, small logic blocks, or an 
entire CPU on a chip?

> The 3090 series machines with very high density
> circuits until the mid 90s.

I had no idea people were still making mainframes in the 90s.

> The s/390 series user cmos technology and were air
> cooled.  The older machines used bipolar technology and many were water cooled.
> Finally the z/series, again cmos technology.

Nice consistent naming scheme. Heh. :-/

> Speeds ranged from .2 MIPS for a
> 360/50, 1 MIP for a 370/158, 15 MIPS for a 3083. The current machines I work
> with today, 2 z/9 BC processors are rated at 600 and 335 MIPS.
> The top of the line z/10 is almost 30000 MIPS.

Now if I could figure out what a typical MIPS rating for a normal PC 
today is, I'd have something dissimilar to compare it to. :-}

Post a reply to this message