A state-of-the-art computer for mission critical tasks. In the "ancient" mid-1960s, all computers were mainframes, since the term referred to the main CPU cabinet. Today, it refers to a class of ultra-reliable medium and large-scale servers designed for enterprise-class and carrier-class operations.
The first mainframe vendors were Burroughs, Control Data, GE, Honeywell, IBM, NCR, RCA and Univac, otherwise known as "IBM and the Seven Dwarfs." After GE and RCA's computer divisions were absorbed by Honeywell and Univac respectively, the mainframers were known as "IBM and the BUNCH."
The Vendors
For decades, IBM has been the dominant vendor in the mainframe business. Although many tried to compete by offering IBM-compatible mainframes, only Amdahl (Fujitsu) has remained as a competitor in this arena (see IBM-compatible mainframe). Unisys, Sun and others also make mainframe-class machines that typically run under a version of Unix or Linux.
There Is a Difference
One might wonder why mainframes cost a million dollars or more when the raw gigahertz (GHz) rating of their CPUs is not any higher than a PC costing 1,000 times less. Quite often in fact, the ratings are lower. Here are the reasons.
Lots of Processors, Memory and Channels
Mainframes support symmetric multiprocessing (SMP) with several dozen central processors in one system. They are highly scalable. CPUs can be added to a system, and systems can be added in clusters. Built with multiple ports into high-speed caches and main memory, a mainframe can address thousands of gigabytes of RAM. They connect to high-speed disk subsystems that can hold terabytes of data.
Enormous Throughput
A mainframe provides exceptional throughput by offloading its input/output processing to a peripheral channel, which is a computer itself. Mainframes can support hundreds of channels, and additional processors may act as I/O traffic cops that handle exceptions (channel busy, channel failure, etc.).
All these subsystems handle the transaction overhead, freeing the CPU to do real "data processing" such as computing balances in customer records and subtracting amounts from inventories, the purpose of the computer in the first place.
Super Reliable
Mainframe operating systems are generally rock solid because a lot of circuitry is designed to detect and correct errors. Every subsystem may be continuously monitored for potential failure, in some cases even triggering a list of parts to be replaced at the next scheduled maintenance. As a result, mainframes are incredibly reliable with mean time between failure (MTBF) up to 20 years!
Here to Stay
Once upon a time, mainframes meant "complicated" and required the most programming and operations expertise. Today, networks of desktop clients and servers are just as complex, if not more so. Large enterprises have their hands full supporting thousands of PCs along with Windows, Unix and Linux and maybe some NetWare and Macintosh for good measure.
With trillions of dollars worth of mainframe applications in place, mainframes may hang around for quite a while. Some even predict they are the wave of the future! See midrange computer.
Mainframe System
UNIVAC Mainframe
Mainframes have provided the computing power for major corporations for more than 40 years. Sperry Rand (Univac), IBM, GE, RCA, NCR, Burroughs, Honeywell and Control Data were the first companies that made mainframes in the U.S. This picture was taken in the mid-1970s. (Image courtesy of Unisys Corporation.)
The first mainframe vendors were Burroughs, Control Data, GE, Honeywell, IBM, NCR, RCA and Univac, otherwise known as "IBM and the Seven Dwarfs." After GE and RCA's computer divisions were absorbed by Honeywell and Univac respectively, the mainframers were known as "IBM and the BUNCH."
The Vendors
For decades, IBM has been the dominant vendor in the mainframe business. Although many tried to compete by offering IBM-compatible mainframes, only Amdahl (Fujitsu) has remained as a competitor in this arena (see IBM-compatible mainframe). Unisys, Sun and others also make mainframe-class machines that typically run under a version of Unix or Linux.
There Is a Difference
One might wonder why mainframes cost a million dollars or more when the raw gigahertz (GHz) rating of their CPUs is not any higher than a PC costing 1,000 times less. Quite often in fact, the ratings are lower. Here are the reasons.
Lots of Processors, Memory and Channels
Mainframes support symmetric multiprocessing (SMP) with several dozen central processors in one system. They are highly scalable. CPUs can be added to a system, and systems can be added in clusters. Built with multiple ports into high-speed caches and main memory, a mainframe can address thousands of gigabytes of RAM. They connect to high-speed disk subsystems that can hold terabytes of data.
Enormous Throughput
A mainframe provides exceptional throughput by offloading its input/output processing to a peripheral channel, which is a computer itself. Mainframes can support hundreds of channels, and additional processors may act as I/O traffic cops that handle exceptions (channel busy, channel failure, etc.).
All these subsystems handle the transaction overhead, freeing the CPU to do real "data processing" such as computing balances in customer records and subtracting amounts from inventories, the purpose of the computer in the first place.
Super Reliable
Mainframe operating systems are generally rock solid because a lot of circuitry is designed to detect and correct errors. Every subsystem may be continuously monitored for potential failure, in some cases even triggering a list of parts to be replaced at the next scheduled maintenance. As a result, mainframes are incredibly reliable with mean time between failure (MTBF) up to 20 years!
Here to Stay
Once upon a time, mainframes meant "complicated" and required the most programming and operations expertise. Today, networks of desktop clients and servers are just as complex, if not more so. Large enterprises have their hands full supporting thousands of PCs along with Windows, Unix and Linux and maybe some NetWare and Macintosh for good measure.
With trillions of dollars worth of mainframe applications in place, mainframes may hang around for quite a while. Some even predict they are the wave of the future! See midrange computer.
Mainframe System
UNIVAC Mainframe
Mainframes have provided the computing power for major corporations for more than 40 years. Sperry Rand (Univac), IBM, GE, RCA, NCR, Burroughs, Honeywell and Control Data were the first companies that made mainframes in the U.S. This picture was taken in the mid-1970s. (Image courtesy of Unisys Corporation.)
 | Reproduced with permission from Computer Desktop Encyclopedia. Copyright (c) 1981-2008 The Computer Language Company Inc. All rights reserved. |
Additional Resources
- From Chapter Three: The Windows Culture
- The pivotal machine in the emergence of the Wintel culture was probably the IBM PC/AT. It sold in sufficiently large quantities to IBM customers to create its own software markets, and was sufficiently poorly engineered and over priced to ensure that its buyers didn't perceive it as a threat to...
- Blog posts 2008-11-14
- Blackbox, SWaP, and the z10
- Blackbox, SWaP, and the z10Over a barrelis where IBM has you if you still use mainframes. Not only is the hardware very expensive, but the labor is too. Universities stopped their mainframes ages ago, so no CS student today has any idea what a mainframe is. This means that you...
- Discussion threads 2008-11-06
- From Chapter Three: The Windows Culture
- From Chapter Three: The Windows CultureOne application per serverThis model is pathetic. Mainframes can run more than one application at a time. So can UNIX machines. Windoze CAN run more than one application at a time, but the more apps you run simultaneously - the higher the risk of failure.This...
- Discussion threads 2008-10-31
- The 'real' cost of application outages
- The 'real' cost of application outagesavailabilityDowntime costs will naturally depend on the business. I worked many years at NYSE, where downtime could cost millions. I used to maintain and run one of their most critical systems. Over 7 years, the total 'unavailable' time was less than 50 seconds.Now working in...
- Discussion threads 2008-10-29
- DynamicOps self service management for virtual environments
- A few days ago, I had the opportunity to speak with Rich Bourdeau DynamicOps' VP Marketing, and learn about Virtual Resource Manager VRM and this spin-off from Credit Suisse. DynamicOps is the result of project to allow virtual resources to be easily configured and provisioned based upon the organization's policies...
- Blog posts 2008-10-29
- Fun with IBM's Z10 numbers
- Fun with IBM's Z10 numbersI just don't believe the IBM numbersin terms of real-world workloads and comparing Linux-to-Linux implementations. Maybe if you are doing some kind of multi-dimensional modeling you might get some advantage from the z10, but for work-a-day web hosting?RE: Fun with IBM's Z10 numbersSo... we're looking at...
- Discussion threads 2008-10-29
- Fun with IBM's Z10 numbers
- As far as I'm concerned, when an organization buys an IBM mainframe to run Linux somebody should be fired - but not the DP guy who bought it: his boss: the non technical guy who approved it without carefully reviewing the alternatives. by Paul Murphy
- Blog posts 2008-10-29
- Open source protectionism?
- I just caught a rather interesting piece by fellow blogger Dana Blankenhorn on the subject of the "value and values" found in open source ecosystems. Near the end, he references an article on InfoWorld which questions whether it is wise to "protect" open source development by mandating its use at...
- Blog posts 2008-10-13
- Easing the SaaS-to-cash cycle
- Easing the SaaS-to-cash cycleCredit crunchMonetization of services becomes all the more critical in these times. As volumes increase operational efficiency becomes critical. At eVapt we therefore go beyond allowing service providers to plug in eVapt billing and payment capabilities. We allow users to scale and automate collection and analysis...
- Discussion threads 2008-10-09
- Data Processing scores big success! Linux loses, again!
- Data Processing scores big success! Linux loses, again!Wrong title, MurphIt should be "IT N00bs at large in Canada" or something along those lines :)Also funny they switched to a windows server. "Oh, we understand windows, cause we use it at home! let's try that! Oh no use! Oh no, what...
- Discussion threads 2008-10-06
- From Chapter Two: The appliance computing culture
- When you look at staffing for the typical iSeries based data center the one thing that's most striking is that relative to the mainframe data center positions are broader, people less constrained, and whole customer facing versus internal distinction is completely missing. by Paul Murphy
- Blog posts 2008-10-03
- Stallman vs. the cloud computing tidal wave
- Stallman vs. the cloud computing tidal waveStallman sounds panicIf FOSS folks believe they can compete with proprietary software, why not build their own FOSS cloud and then take the battle right to the other proprietary cloud platforms?Ever been in a restaurant kitchen ?How old is the food ? Where have...
- Discussion threads 2008-10-01
- Thinking about Oracle Fusion applications
- Thinking about Oracle Fusion applicationsKilling Oracle DB in the enterprise......will probably be more difficult that killing off the mainframe...Postgres and/or MySQL may match, or even pass, it in terms of capabilities. But Oracle will still have history.And remember, Oracle pretty much works. This isn't like replacing Windows, where...
- Discussion threads 2008-09-30
- Oracle enters hardware market; Launches storage server to ride shotgun with database
- Oracle enters hardware market; Launches storage server to ride shotgun with databaseRE: Oracle enters hardware market; Launches storage server to ride shotgun with databaseIt's difficult to see any advantage over what I could buy from HP or any of the major vendors. It appears to be an HP box...
- Discussion threads 2008-09-25
- Hey, Microsoft. About those ads... Yeah, like that.
- Hey, Microsoft. About those ads... Yeah, like that.In fairness......no-one likes you at all.re: I WAS a PC[i]I WAS a PC... And I am in my daily life. Got tired of the constant care, feeding, and BS associated with it.Now I'm a Mac and make my living fixing the PC mess.[/i]Ohh,...
- Discussion threads 2008-09-19
- Has IBM Figured Out The Art Of The Sale Better Than Bill
- Has IBM Figured Out The Art Of The Sale Better Than BillJust like their Linux ads?That answered little more then nothing.Are they really all that savey to the medium, themselves?Comparative savvinessYah, it's certainly no slam-dunk that IBM's attempt at tongue-in-cheek humor worked or works. Doubt it could say that the...
- Discussion threads 2008-09-18
- Has IBM Figured Out The Art Of The Sale Better Than Bill & Jerry?
- So the broadcast video ads with Bill Gates and Jerry Seinfeld for Microsoft are over for now. Maybe you're among those who disliked the ads because they never answered the simplest of fundamentals of an a: What are these guys selling? So traipse...
- Blog posts 2008-09-18
- VMworld - Dr. Stephen Herrod's keynote
- 20080917 VMworld Dr Stephen Herrod Keynote I'm not sure why conference directors think that people enjoy walking into a darkened room, facing disorienting flashing lights coming from the front of the room, watching company ads and customer profiles on overhead videos combined with extremely loud, unpleasant, pulsing...
- Blog posts 2008-09-17
- Economic crisis as a technology change agent
- Sometimes, to make an omelette, you need to break a few eggs. However if this week's bloodbath on Wall Street is any indication, then I would say we've made ourselves a full-blown deep dish quiche the size of Yankee Stadium with tens of billions of cartons...
- Blog posts 2008-09-16
- NEC tees up six-core Intel Dunnington server
- NEC on Tuesday will announce one of the first servers built from the ground up to take advantage of Intel's six-core Dunnington chip. NEC, which is looking to make a splash in the U.S. market, is focusing its efforts on a new server line designed for Intel's...
- Blog posts 2008-09-15
Neighboring Terms
Premier Vendor Content Whitepapers, webcasts & resources from our Power Center Sponsors
-
- Visit the ZDNet Enterprise Mobility Hub sponsored by Verizon Wireless! It has the latest news, commentary, and resources for your mobile and wireless strategy development.
- Video: How to spoof a MAC address
- Five steps to protect mobile devices anywhere, anytime
- Fetch Your E-Mail from Anywhere
- From our sponsors
-
- Push to Talk from Verizon Wireless Choose Verizon Wireless and get a Push to Talk service you can rely on. The only service that comes with the Network. Starting at $699.
-
