The end of an era

[BobSch]

NASA is pulling the plug on their last mainframe.

http://www.techrepublic.com/blog/tec...68?tag=nl.e101

[pelligrini]

They probably had to, because the guys with the slide rules that put it togeather and kept it running have retired.

[jackellis]

They probably had to, because the guys with the slide rules that put it togeather and kept it running have retired.

Hey, watch it! I still have a half dozen slide rules and I still use them from time to time.

Calculators and spreadsheets are for weenies.

Real men (and women) do the figures in their heads. A lost art.

[cabinetman]

Does anybody remember the IBM 1401?

.

[Denco]

I had a slide rule, but it slid off the table and broke. It's last words were, "This doesn't add up".

[LCHIEN]

i7 with six 64-bit cores, with 1.17 billion transistors...

That chip can do 158.4 GFLOPS - 158 x 10^9 floating point operations per second. Half of that in double precision (according to this doc: http://www.intel.com/standards/floatingpoint.pdf)

If someone can figure out from the ibm website what any model of the Z9 can do GFlop wise, my hat's off to you.

[unknown poster]

Try searching for IBM mainframe specs in MIPs, not GFLOPS. I guess IBM assumes you won't be using those machines just for floating point operations so they spec them in instructions per sec rather than floating point operations per sec.

[guycox]

Does anybody remember the IBM 1401?

.

I was of the 1620 persuasion.

[jackellis]

My wife was a systems programmer at IBM back in the 1960s. IBM trained her because there were no computer science programs in those days, and she did her coding in assembly language. Her first assignment was writing what we know call device drivers for an early IBM disk drive. Later on she ported the Berkeley Unix Assembler to IBM's first RISC workstation (preceded the Power line). She still has her green plastic card.

Today's programmers are spoiled (I know, spoken like a true Luddite). The first computer I ever worked on was a Univac 1108, complete with card reader and something like a half megabyte of magnetic core memory. Later on I got to do some programming on what by then had become an ancient Burroughs machine with 32k of memory. The debugging tools and other aids we take for granted today did not exist. When I first entered the working world as a co-op student and did some programming, debugging code sometimes required reading memory dumps. Since I was working at a remote site, I time-shared using a teletype machine and a 300 baud phone connection. I dont miss those days.

[jackellis]

Postscript.

I sent my wife the link, who told the story of a colleague at IBM at the time who was from Shanghai and would do an occasional bit of hexadecimal arithmetic out loud in Mandarin.