[email protected] (Dana Mitchell) writes:
> Exactly!  They are just managing the decline and extracting maximum
> profit out of it along the way, IBM (and more importantly Wall St.)
> have no interest in expanding the z business.
>
> I wouldn't exactly consider cloud a high margin business.

there was scenario in the 90s when wallstreet probably cared.

1992 IBM had gone into the red and was being reorganized into the
13 "baby blues" in preparation for breacking up the company.

At the time, large number of customers were moving off mainframes to
other platforms (major factor in company going into the red) ...
however large part of the financial industry & wallstreet was still
heavily invested in mainframes (I've periodically told story about
senior disk engineer doing talk in the late 80s at annual, world-wide,
internal communication group ... claiming the communication group was
going to be responsible for demise of disk division, they were seeing
data fleeing mainframe datacenters with drop in disk sales).

The former president of AMEX .... which had been in battle with
KKR for private-equity take-over of RJR ... and KKR won:
https://en.wikipedia.org/wiki/Barbarians_at_the_Gate:_The_Fall_of_RJR_Nabisco

KKR ran into problem with RJR and hired away president of AMEX to turn
it around. Then the IBM board hired him away to resurrect IBM and
reverse the breakup ... using some of the same techniques that were
used at RJR:
http://www.ibmemployee.com/RetirementHeist.shtml

Also in 1992, AMEX spun off a lot of its dataprocessing into independent
business unit which was the largest IPO up until that time.  Besides
doing a lot of AMEX stuff, it also does a significant amount of
transaction, statementing, account management, etc outsourcing for large
number of other financial institutions (at one point handling everything
for over half of all credit card accounts in the US).  It has been one
of the largest mainframe customers (total trivia, 15yrs later, KKR does
private-equity take-over of that business, in the largest reverse-IPO up
until that time).

Starting in the middle 90s, major financial institutions spent billions
of dollars to move off mainframe cobol overnight batch settlement to
straight-though processing on large number of "killer micros". Over
decades, batch financial had been front-ended with "real-time"
transaction, but they were really just being queued for overnight batch
processing window. In the 90s, globalization was currenting size of
overnight batch window and singificantly increasing workload
... breaking the overrnight window. Turns out the efforts were using
some standard parallization libraries that had 100 times the overhead of
cobol batch. The toy demos looked good ... but pointing out to them that
they wouldn't scale was ignored. They had to go into deployments that
went down in spectacular flames (100 times overhead totally swamping any
throughput anticipated with large number of "killer micros").

Middle of last decade I was involved in project that did high-level
specification of business processes that was then decomposed into large
number of light-weight SQL statements.  It leveraged the significant
efforts by major RDBMS vendors (including IBM) in enormous throughput
for parallized cluster operation. It easily demonstrated the required
throughput for straight-through processing along with being able to
easily add all sort of enhancements (helped by advances in "killer
micros" gave them throughput compareable to mainframes). Did well
accepted demonstrations for major financial associations ... and then
hit brickwall. Was finally told that there were too many executives that
bore the scars of the failed efforts in the 90s and it was going to have
to wait for new generation of executives (which could be demise of one
of the last major mainframe cash cows).

Trivia: I was involved in the original relational/SQL implementation,
System/R and tech transfer to Endicott for product release. At the time
this was done "under the radar" because most of the company was focused
on the strategic next generation DBMS, (code-name) EAGLE. When EAGLE
imploded, there was request for how fast could System/R be available on
MVS ...  which was eventually released as DB2 (originally for decision
support only).

-- 
virtualization experience starting Jan1968, online at home since Mar1970

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to