Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : IBM -- Ignore unavailable to you. Want to Upgrade?

To: VICTORIA GATE, MD who wrote (5106)	5/2/1999 7:12:00 PM
From: J R KARY	Read Replies (1) \| Respond to of 8218

Ahead of plan and no fear of the sky falling in IBM's G 6 announcement Not a mention of Y2K , or even Jules' puts , instead forecasting IBM's new G 6 will add to the G 5's server success , which settled for a humble 95 % of market share: -------- " Mainframes, despite IBM's diversification into software and computer services, remain key to financial performance for IBM, the world's largest computer maker... ...IBM hopes the G6 machines will capitalize on the momentum of the G5 Server, which analysts say has captured as much as 95 percent of worldwide sales of mainframe computers over the last three quarters, leaving only crumbs for Hitachi Data and Fujitsu Ltd.'s Amdahl, IBM's two direct competitors. -------- 95 % is that all we can expect from IBM ? Thimk again: ------------ "Each mainframe sale tends to bring in additional software and services revenue. ''Obviously, the more successful we are shipping 390 (computer) capacity, the more it benefits our middleware (software) business, our financing business, and to some degree our services business,'' (IBM's )Carlucci said. -------- Ships in May so the G 6's revenues will show up in 2Q99 -- Let the good times roll ! Jim K.

To: VICTORIA GATE, MD who wrote (5106)	5/3/1999 12:45:00 PM
From: Arrow Hd.	Read Replies (2) \| Respond to of 8218

Victoria, thanks for posting this great article. This is must reading for anyone who truly wants to get a general education of what the mainframe is all about and why customers use them. I would have been proud to have penned this article myself (no, it is not mine). There is one point they tried to make but unless you knew the architecture it was too general to pick up and understand. Where they mentioned that the IBM Mainframe can allocate 100% of its capacity versus the other platforms such as HP or SUN where you need additional standby machines, this is what they were alluding to: The IBM mainframe is manufactured with a full engine compliment within the machine. The MCM (multi chip module) is a field replaceable part that contains all of the engines. Each engine with G6 is a 200 plus MIP processor that can work in conjunction with the other engines comprising a huge single image to process major applications such as business intelligence or data mining or the engines can be logically partitioned to run smaller jobs independently of other jobs running on other engines within the MCM. Not quite Nway versatility but very close. "Allocating 100% capacity" means that if a customer needs only five active engines to process their business with say two additional engines to be allocated to one SAP (runs I/O function) and one more to run an ICF (integrated coupling facility for a parallel sysplex where not just engines share job routines but machines share with each other) then the customer has used up seven engines but the MCM has 12 engines available. So how are these five "extra" engines used? First, if one of the active seven fail one of the extra or spare engines is invoked and with G6 this is non-disruptive without an IML to reload microcode. Repairs itself automatically. Now lets say this customer is a retailer who has just gone with IBM's E-business solution and it really takes off and it is also December and the Xmas season crush has them out of machine capacity and the factory is sold out due to the fourth quarter push, what do we do here with this dilemma? Its called capacity on demand! The spare engines, which the customer did not have to pay for with the initial machine purchase, unlike the HP or Sun solution where they had to buy the extra spare machines that are sitting idle, can be activated by the down loading of machine microcode, called Licensed Internal Code and in special situations where specific contracts are in place, that code already resides on the machine and it is activated with asset protected passwords. So it is Xmas eve and the last shipment is out but now they have to do their accounts receiveable application and they just dont have the capactiy. They immediately place an order with IBM for lets say two additional engines for another 420 MIPs of processing power. The IBM factory, which operates 24 by 7 gets the order through the emergency hotline from the IBM account team, manufacturing cuts the new load of code, downloads it machine to machine and activates in a non-disruptive process the two new engines and the data center immediately has a machine with an extra 420 MIPs and still three spare engines to use however needed. Oops, another machine in the data center just died (probably an Hitachi or Amdahl machine) and we need backup! This customer planned ahead by contracting with IBM for IBM's emergency backup feature for LIC controlled processors. So, without hesitation or missing a cylce another call to the hotline invoking the EBU feature and within minutes the IBM team is on a roll again downloading more LIC code to invoke the remaining three engines on our initial machine offloading job streams to keep this customer running. If they had the parallel sysplex architecture then we not only have additional engines to invoke but additional machines with their additional engines for Nway operations. The retailer down the street who has the Unix implementation is still on the phone tracking down the HP or Sun customer engineer to get out of bed to come to their data center but he swears his car battery is dead and he wont be able to get there until the morning when he can call the local garage station for a new battery. For those who still think that the decentralized approach of Wintel, Unix or some other solution is the way to go as we enter the 21st century then you have missed a major architectural paradigm shift. Capacity on demand allows customers to pay for what they need, have custom developed contracts for future capacity requirements and get it within hours, non-disruptively in an almost Nway performance model.