On the mainframe’s 50th birthday, IBM positions it because the still young Enterprise Cloud System, able to running 6,000 Linux workloads.

An IBM mainframe will also be subdivided to run Linux virtual machines, in something of an analogous manner that Amazon Web Services or the Rackspace cloud runs virtual machines. But unlike an AWS host, a single mainframe runs 6,000 VMs at a time.

Likewise, the old mainframe Customer Information Control System, better called CICS, can process 1.1 million transactions per second. It was the predecessor to client/server architectures and connected thousands of users to a single mainframe. That number, incidentally, is more transactions than are conducted by the Google search engine per second, based on a spokesman for IBM partner GT Software in Atlanta.

On the mainframe’s 50th anniversary this week, IBM buffed up and displayed it as probably the most few systems which could scale to cloud-like proportions. It could still stay alongside of the x86 server clusters that power most of the Web-scale companies which have emerged within the last 10 years. IBM claimed the mainframe is an appropriate platform on which to run the enterprise private cloud, and says public cloud service providers can adopt it for Linux workloads in addition. The mainframe, as IBM would have it, is calling young again.

Cloud architectures often seem like an try to solve with commodity parts the similar problems that were answered by the 1st mainframe’s design — general-purpose operations combined with an enormous degree of scalability. The mainframe brought concentrated CPUs and memory together, then added access to communications and storage via nearby, high-speed channels. It was an early model of intense computation. And it’s enabled IBM, for a lot of of the past 50 years, to maintain mainframe system prices high. Within the cloud, compared, low-cost pc processors are packaged with nearby disks and switches on a server rack to provide one of the crucial same attributes. Google, Amazon Web Services, and eBay have all been built on those commodity parts.

[Wish to learn more about IBM’s $1.2 billion investment in cloud infrastructure? See IBM Preps SoftLayer Cloud Datacenter In Hong Kong.]

Of late, IBM have been lowering mainframe prices in order that it may possibly continue to play a job within the cloud. On Tuesday, it took steps to make the mainframe a more attractive host for cloud service suppliers. It announced the primary cloud-oriented, System z-based offering, the IBM Enterprise Cloud System. The Cloud System relies on a zBC12 or zEC12 mainframe. Because it was announced last July, the zBC12 retailed for approximately $75,000. Compared, 2003’s mainframe model, the z990 T-Rex, retailed for $1 million.

In its announcement, IBM said the mainframe is now a less expensive environment than x86 servers on which to run Linux virtual machines. Unlike the mainframes that run IBM’s proprietary operating systems, the Enterprise Cloud System is geared to run Linux and Linux virtual machines. “Due to higher system efficiency and larger scalability, the full cost of a few Linux on System z cloud deployments is also as much as 55% not up to comparable x86-based cloud infrastructure,” IBM claims.

The cloud mainframes are equipped with special power processors geared to running Linux virtual machines, called Integrated Facility for Linux (IFL). Each IFL can host 60 virtual machines, and a zEC12 is able to mounting 100 IFLs. Hence, IBM comes up with a figure of 6,000 VMs for a single mainframe host.

An Enterprise Cloud System, as well as a mainframe Linux server, includes IBM v7000 or DS8000 storage; IBM Wave z/VM (a graphical management interface for managing VMs); and IBM Cloud Management Suite. The last includes SmartCloud Orchestrator for configuring and deploying virtual machines; Omegamon XE for monitoring and managing performance of workloads; and Tivoli Storage Manager. The suite is the IBM software that gives the automatic Linux VM spinup, deployment, and management.

Also on Tuesday, IBM introduced new consumption-based pricing models for managed enterprises. If a service provider builds infrastructure in response to mainframes, it could repay the mainframe bill according to its use by customers, rather than an upfront payment.

IBM’s example of an organization that’s doing so is Business Connexion, the biggest enterprise company in Africa. The firm is packaging mainframes into “pop-up” datacenters that may be installed in a telco’s remote office to supply Internet services to a previously unreached area. A mainframe during this setting uses concerning the same amount of electricity as a garments dryer, IBM spokesmen said.

Private clouds are moving rapidly from concept to production. But some fears about expertise and integration still linger. Also within the Private Clouds Step Up issue of InformationWeek: The general public cloud and the steam engine have more in common than you could think. (Free registration required.)

Charles Babcock is an editor-at-large for InformationWeek, having joined the publication in 2003. He’s the previous editor-in-chief of Digital News, former software editor of Computerworld and previous technology editor of Interactive Week. He’s a graduate of Syracuse … View Full Bio

More Insights