IBM introduces a new mainframe computer targeted to some markets better served by clusters of smaller servers

2-Color-Design-Hi-Res-100px-widthIBM debuted its z Systems line of mainframe computers in January, 2015. The product line is targeted to a hot segment of enterprise computing consumers, organizations with a burning need to manage mobile devices, users availing of cloud computing offers and, above all, secure online data processing.

Readers can learn more about this product line on its website, IBM z Systems (http://www-03 NULL.ibm NULL.com/systems/z/announcement NULL.html?lnk=ushpls1). A quick glance at the marketing communications content reveals some popular and absolutely current computing themes:

  • mobile computing
  • enterprise social computing
  • real-time analytics and “in-transaction” analytics
  • secure, cloud computing

Video product presentations are available on the product website, along with a traditional datasheet in PDF format. A quick glance at the datasheet exposes a cluster approach to delivering the computing power of a typical high performance computing (HPC) system. No problem so far, but how does the introduction of a hardware computing platform with these capabilities align alongside IBM’s announced effort to become a major player in the public and private cloud market for IaaS, PaaS and SaaS? Does it make sense for developers building solutions for Hadoop, Map Reduce, and other cluster architectures optimized for lots of comparatively much smaller CPUs to focus on porting these applications over to IBM’s platform?

Notable on the datasheet is IBM’s suggestion about development platforms. The recommendation is for Java. But a lot of the most promising sector of app development, in, admittedly, very early 2015, is built on scripting languages, with JavaScript getting the most attention. There is also some substantial mention of traditional mainframe computing languages (COBOL) on the datasheet.

So one needs to question just how this product line adds value to IBM’s effort to catch up with its peers in the cloud computing business with systems like these. Certainly a review of the website for these new products is recommended given the themes articulated by the press releases about the z System computers. The most prominent of these are all allusions to mainframe computing, which, in 2014, seems to be something of anathema. Big iron is, unfortunately, no longer the recommended way for most enterprise businesses to proceed, at least not for the ones already committed to Azure, AWS and Google Compute Engine.

Ira Michael Blonder

© IMB Enterprises, Inc. & Ira Michael Blonder, 2015 All Rights Reserved


Are Valuations of Cloud IaaS Offers Too High?

Are valuations of Amazon, Google and even Microsoft too high as the result of overly optimistic revenue growth assumptions for GCE, AWS and Azure? A brief look at why the value of any one of these offers may elude SMBs, reveals why inflated expectations may be at play here.

Reason #1: Marketing Collateral is Opaque for Average Business Users

It’s not likely a CIO from a $100 Million SMB is going to make much of the information Amazon provides on its Amazon EC2 Instances (http://aws NULL.amazon NULL.com/ec2/instance-types/) web page. When the sheer opacity of this promotional information is put together with a complete absence of easily accessible human support (I challenge the reader to find a telephone number for questions on any of the pages of the presentation), the inescapable conclusion is the target market for this type of an offer is a very small set of very large organizations in the public, private, and not for profit sectors.

How much of this very small market (composed, admittedly, of consumers with very large internal appetites) will any one of these competitors seize, and for how long? I argue all three services are completely targeted at the very same market.

Reason #2: Monthly Costs Are Too High for SMBs. Better Offers Are Available Elsewhere

For an average $100 Million SMB looking to use one of these three offers to provide a flexible, virtualized server platform for a big Microsoft® application like SharePoint® 2013, on premise, the monthly costs are simply too high. Amazon AWS, per a look at their publicly posted pricing schedule for IaaS compute instances on April 10, 2014, charges just about $1.28 per hour for the horsepower required to drive server, Search services, SQL services, etc. For 7×24 up time, we’re looking at $928.00 per month operating costs (this estimate is based on AWS’ “Windows with SQL Standard” offer and an m3.xlarge instance, which includes 4 CPUs and 13GBs of RAM).

It won’t take many months of this type of cost for the expense of opting for this support method to exceed the cost of purchasing an HP Blade Server with comparable specifications for physical support of the application on premise. Better yet, this same SMB can purchase vSphere from VMware and run its own internal private cloud IaaS off of the HP Blade Hardware.


Leading competitors for cloud IaaS offers are still focused on the very top of the market, without really viable offers for the vast market opportunity below. Unless/until these offers emerge, on premise alternatives still look like the preferred approach for SMBs.

Disclaimer: I’m long Microsoft, have no position in Amazon, Google or VMware.

Ira Michael Blonder (https://plus NULL.google NULL.com/108970003169613491972/posts?tab=XX?rel=author)

© IMB Enterprises, Inc. & Ira Michael Blonder, 2014 All Rights Reserved


Is Google’s Compute Engine (GCE) a Direct Competitor to AWS or Microsoft Azure?

Google publicly introduced its Google Compute Engine (GCE) (https://cloud NULL.google NULL.com/products/compute-engine/) as an actual product back in December 2013. A lot has been written about GCE since its debut, with specific reference to this product as a direct competitor to Amazon AWS. But the lack of support for any Windows O/Ss, databases, etc (at least as far as I could find) says they are in a market of their own.

Virtual machine offers for operating systems are limited to two Linux distributions: Debian, and CentOS. Where Amazon offers RDS for managing Oracle, and even SQL Server in the cloud, GCE does not offer a competitive solution. Neither has GCE magnetized a level of support from the developer community to successfully compete with AWS. Quite a number of third party Apps are available for AWS, including quite a few useful for managing SAP database products.

So what should anyone following the cloud market for IaaS and SaaS solutions make of the dramatic price reductions Google recently announced for GCE. I think it’s a mistake to plan on a lot of pressure on AWS from these price reductions. I think it makes more sense to read the price discounting as emblematic of difficulties Google has experienced spreading the word on GCE and capturing a sufficient share of the market for these services. Perhaps it’s safe to say Google GCE has simply been successful magnetizing interest from the Open Source community, all the way from early stage ISVs to larger organizations like the Institute for Systems Biology portrayed in a case study on the GCE web site.

In contrast, AWS has been adopted by organizations, here in the US, of considerable size, not the least of which is the Federal CIA. A lot of this success is likely attributable to the availability, through AWS, of support for databases, and related applications, written for Windows computing environments.

It’s likely Google will expand the list of GCE available VMs to include support for the Microsoft Windows computing paradigm sometime soon. But in the meantime, Azure looks a lot more like a direct competitor to AWS than does Google GCE

Ira Michael Blonder (https://plus NULL.google NULL.com/108970003169613491972/posts?tab=XX?rel=author)

© IMB Enterprises, Inc. & Ira Michael Blonder, 2014 All Rights Reserved