Mary, >>1. What is the prognosis for System 390 type of computing?
To the thread, this post is long and is mostly about mainframes (off topic).
By prognosis, if you mean S390's health going forward, excellent, for all of the reasons I put into my previous post.
>>2. Do they provide for visual computing (by themselves without going through front-end
manipulation and or exotic emulations)?
No, for visual computing, people would use networked front ends, like PC's or workstation, that ultimately "report" to the mainframe.
>>3. Are current System 390 sales mostly upgrades; and, or, what is the rate of sales of System 390s
for brand new applications or new organizations?
Right now, most S390 revenues (by a large amount) are new machine sales. This is because, as I said in my previous post, the CMOS based machines have surpassed the old ECL ones in terms of performance, CPU for CPU, just early this year, in terms of performance. Now, customers will swap out their old ones on an engine for engine basis, e.g., remove a 12 way ECL and put in a 12 way CMOS. The last post gives the advantages of doing this (getting rid of ECL and putting in CMOS).
>>4. Are current System 390 type applications mostly stuff that were implemented earlier, such as
for Airline Reservations (United, AMR, DAL);Car Rental (Hertz, Avis, National, etc); Banking going through front-end
manipulation and or exotic emulations)?
All of the above and a lot of new applications. Some of the new ones are at companies that tried to use strings of PC's, workstations and networks in the early 90's for enterprise computing. Some of them failed and they went to mainframes.
>>I am assuming that these large systems are mostly character based and that they were developed
years ago and has over the years been upgraded and enhanced.
Yes. Technology has taken them up in performance at a rate even faster than with microprocessors, IMO. Part of this is that IBM still leads in IC technology and its enhancement rate.
>>If my assumptions are correct, how would you characterize these systems in terms of what these
companies think of them?
Not sure I understand this one. If it's what do companies that use mainframes think of mainframes, they must like them, because there is a lot of repeat business happening.
>>Do most end users feel that these systems are antiquated, not responsive to their real needs, and
transactions (CITI, BankAmerica, Wells Fargo...)?
No. Because of performance, scalability, RAS, I/O bandwidth, security, etc., etc., mainframes are still considered it for enterprise computing at Fortune 100, Fortune 500 companies, etc.
Re: "My main reason for asking these questions is to gauge the possibilities and time table for Merced
to penetrate this market (in 5 to 10 year time span) - if that is a possiblity at all."
Anything is possible. But, even as Merced(es) speeds ahead, so will the mainframes.
Tony |
