PureSystems from IBM are the latest converged data center hardware, systems which pack compute, storage and networking into one enclosure, to hit the market. Although a bit late to market these systems have some interesting differentiators and are a good fit for organizations that have an investment in IBM technologies.
PureSystems splits the difference between single-purpose appliances a la Netezza, Exadata and others in the markt and general-purpose hardware. Cisco UCS publicly claims that they have been able to walk that fine line between tight integration and flexibility and we think IBM has developed an approach that will allow the same boast.
Mix-and-match compute nodes
In IBM’s case, a buyer can mix compute nodes in the same enclosure and run different operating systems - x86, IBM Power, Linux, AIX, IBM i and Windows - in the same enclosure.
These systems, sold under the banner of IBM PureSystems, will run IBM’ SmartCloud, which is a good way for companies to bootstrap a private cloud initiative. The claim (verified) is that with the SmartCloud integration you can basically do a four-click deployment to provision a new cloud — you identify the resources, pool them, pick the applications, then deploy and run.
The scale-up vs. scale-out debate revived
Particularly with the buzz around big data, OpenStack and much of the innovation happening, conventional wisdom holds that the best way to attack cloud computing is with massively scaled-out commodity hardware.
However, in many cases - particularly where an organization may not have the luxury of time - or ready access to the sometimes hard to find talent, PureSystems’ pre-integrated model and flexibility could make a difference. The fact that you can mix and match multiple blades and move the virtual environment from one blade to another if needed, and have it managed automatically by policy is an interesting value proposition. The following video is a time lapse of a 4 hour setup:
The net result, is that PureSystems give users the simplicity of an appliance, the flexibility of the traditional computing model and the automatic scale-up/scale-down of the cloud computing model.
As with most things, the market will tell. In the mean time, watch this space for more details on the engineering we find interesting and what practical experience tells as as we roll these things out to clients.