> IBM’s mainframe monopoly was suddenly
challenged by minicomputers from companies
like DEC, Data General, Wang Laboratories,
Apollo Computer, and Prime Computers.
So, to shed some more light on this
statement, especially about "mainframe
monopoly", let me recount some of my
history with IBM mainframes:
(1) Uh, to help work myself and my wife
through grad school, I had a part time job
in applied math and computing: Our IBM
Mainframe TSO (time-sharing option) bill
was about $80,000 a year, so we got a
Prime, and soon with my other work I was
the system administrator. Soon I
graduated and was a new B-school prof
where the school wanted more in computing.
So, I led an effort to get a Prime -- we
did. IBM and their super-salesman Buck
Rodgers tried hard but lost.
The Prime was easy to run, very useful,
and popular but would not have replaced
IBM mainframe work running CICS, IMS, DB2,
etc. Of course, in a B-school, we wanted
to run word processing, D. Knuth's TeX
math word whacking, SPSS statistics, some
advanced spreadsheet software (with linear
programming optimization), etc. and not
CICS, IMS, DB2.
(2) Later I was at IBM's Watson lab in an
AI group. For our general purpose use,
our lab had six IBM mainframes, IIRC U, V,
W, X, Y, Z. As I recall they had one
processor core each with a processor
clock likely no faster that 153 MHz.
Okay, in comparison, the processor in my
first server in my startup is an AMD
FX-8350 with 8 cores and a standard clock
speed of 4.0 GHz.
So, let's take a ratio:
(8 * 4.0 * 109)/(6 * 153 * 106) = 34.9
so that, first cut, just on processor clock
ticks, the one AMD processor is 35 times
faster than all the general purpose
mainframes at IBM's Watson lab when I was
there.
But, still, on IBM's "mainframe monopoly",
if what you want is really an IBM
mainframe, e.g., to run old software, then
about the only place to get one is from
IBM. So, IBM still has their "mainframe
monopoly".
Or to be extreme, an Apple iPhone, no
matter how fast it is, does not really
threaten the IBM "mainframe monopoly".
Continuing:
> ... like DEC, Data General, Wang
Laboratories, Apollo Computer, and Prime
Computers. And then, scarcely a decade
later, minicomputers were disrupted by
personal computers from companies like
MITS, Apple, Commodore, and Tandy.
Not really: The DEC, DG, ..., Prime
computers were super-mini computers and
were not "disrupted" by the PCs of "MITS,
Apple, Commodore, and Tandy."
The super-mini computers did lose out but
later and to Intel 386, etc. chips with
Windows NT or Linux.
> ... Microsoft the most powerful company
in the industry for two decades.
Hmm. So now Microsoft is not so
"powerful"? Let's see: Google makes it
easy to get data on market capitalization:
Apple: $1,308.15 B
Microsoft: $1,202.15 B
Alphabet: $960.96 B
Amazon: $945.42 B
Facebook: $607.59 B
Exxon-Mobil: $297.40 B
Intel: $256.35 B
Cisco: $201.47 B
Oracle: $173.73
IBM: $118.84 B
GM: $50.22 B
Microsoft is still a very powerful
company.
Uh, I'm no expert on Apple, but it appears
that the Apple products need a lot of
access to servers, and so far they tend to
run on processors from Intel and AMD with
operating system software from Microsoft
or Linux -- that is, Apple is just on the
client and not the server side.
It appears, then, that in computing
Microsoft is the second most powerful
company and is the most powerful on the
server side.
Sure, maybe some low power ARM chips with
3 nm line widths and Linux software will
dominate the server side, but that is in
the future?
And personally, I can't do my work with a
handheld device, need a desktop, and am
using AMD and Microsoft and nothing from
Apple. A Macbook might suffice for my
work but seems to cost maybe $10,000 to
have the power I plugged together in a
mid-tower case for less than $2000.
Broadly it appears that the OP is too
eager to conclude that the older companies
are being disrupted, are shrinking and are
fading, are being replaced, etc.
Maybe the main point is just that in the
US hamburgers were really popular and then
along came pizza. So, pizza is popular,
but so are hamburgers!
There is
> IBM’s mainframe monopoly was suddenly challenged by minicomputers from companies like DEC, Data General, Wang Laboratories, Apollo Computer, and Prime Computers.
So, to shed some more light on this statement, especially about "mainframe monopoly", let me recount some of my history with IBM mainframes:
(1) Uh, to help work myself and my wife through grad school, I had a part time job in applied math and computing: Our IBM Mainframe TSO (time-sharing option) bill was about $80,000 a year, so we got a Prime, and soon with my other work I was the system administrator. Soon I graduated and was a new B-school prof where the school wanted more in computing. So, I led an effort to get a Prime -- we did. IBM and their super-salesman Buck Rodgers tried hard but lost.
The Prime was easy to run, very useful, and popular but would not have replaced IBM mainframe work running CICS, IMS, DB2, etc. Of course, in a B-school, we wanted to run word processing, D. Knuth's TeX math word whacking, SPSS statistics, some advanced spreadsheet software (with linear programming optimization), etc. and not CICS, IMS, DB2.
(2) Later I was at IBM's Watson lab in an AI group. For our general purpose use, our lab had six IBM mainframes, IIRC U, V, W, X, Y, Z. As I recall they had one processor core each with a processor clock likely no faster that 153 MHz.
Okay, in comparison, the processor in my first server in my startup is an AMD FX-8350 with 8 cores and a standard clock speed of 4.0 GHz.
So, let's take a ratio:
(8 * 4.0 * 109)/(6 * 153 * 106) = 34.9
so that, first cut, just on processor clock ticks, the one AMD processor is 35 times faster than all the general purpose mainframes at IBM's Watson lab when I was there.
But, still, on IBM's "mainframe monopoly", if what you want is really an IBM mainframe, e.g., to run old software, then about the only place to get one is from IBM. So, IBM still has their "mainframe monopoly".
Or to be extreme, an Apple iPhone, no matter how fast it is, does not really threaten the IBM "mainframe monopoly".
Continuing:
> ... like DEC, Data General, Wang Laboratories, Apollo Computer, and Prime Computers. And then, scarcely a decade later, minicomputers were disrupted by personal computers from companies like MITS, Apple, Commodore, and Tandy.
Not really: The DEC, DG, ..., Prime computers were super-mini computers and were not "disrupted" by the PCs of "MITS, Apple, Commodore, and Tandy."
The super-mini computers did lose out but later and to Intel 386, etc. chips with Windows NT or Linux.
> ... Microsoft the most powerful company in the industry for two decades.
Hmm. So now Microsoft is not so "powerful"? Let's see: Google makes it easy to get data on market capitalization:
Apple: $1,308.15 B
Microsoft: $1,202.15 B
Alphabet: $960.96 B
Amazon: $945.42 B
Facebook: $607.59 B
Exxon-Mobil: $297.40 B
Intel: $256.35 B
Cisco: $201.47 B
Oracle: $173.73
IBM: $118.84 B
GM: $50.22 B
Microsoft is still a very powerful company.
Uh, I'm no expert on Apple, but it appears that the Apple products need a lot of access to servers, and so far they tend to run on processors from Intel and AMD with operating system software from Microsoft or Linux -- that is, Apple is just on the client and not the server side.
It appears, then, that in computing Microsoft is the second most powerful company and is the most powerful on the server side.
Sure, maybe some low power ARM chips with 3 nm line widths and Linux software will dominate the server side, but that is in the future?
And personally, I can't do my work with a handheld device, need a desktop, and am using AMD and Microsoft and nothing from Apple. A Macbook might suffice for my work but seems to cost maybe $10,000 to have the power I plugged together in a mid-tower case for less than $2000.
Broadly it appears that the OP is too eager to conclude that the older companies are being disrupted, are shrinking and are fading, are being replaced, etc.
Maybe the main point is just that in the US hamburgers were really popular and then along came pizza. So, pizza is popular, but so are hamburgers!
I also go along with the point of zozbot234 at
https://news.ycombinator.com/item?id=21986141
> Software is still eating the world, and there will be plenty to eat for a long time.