Right now I am in China talking with APJ customers as a follow-on to our recent announcement - delivering the first Mission-critical Converged Infrastructure. It's been interesting. I've been asked by several IT managers if this is real. So let me just state this for the record ... HP Converged Infrastructure it is real and can be delivered today! Assuming others have this same question, let me expound a bit. What this announcement does is "bullet proof" your infrastructure by bringing our proven Mission Critical Intellectual Property to the Converged Infrastructure. This 'integration by design' brings the best of both worlds - all based on openness so you get choice and the benefit of industry standards for designing your Data Center of the future. It's not about the chip per se, rather about bringing in Mission Critical across the full technology stack. Let me give you a few examples:
- Mission Critical Flex-Fabric and Always-on Resiliency (video) - The Crossbar Fabric in Superdome 2 eliminates any single point of failure in the data path of systems to give 450% boost in server reliability.
- The Blade Link using Intel's Quick Path Interconnect (QPI) high performance link technology enables you to scale up your Integrity blades from 2 to 4 sockets or 4 to 8 sockets.
- Superdome 2 Analysis Engine provides predictive error detection and recommendations and fixes errors where appropriate.
- HP Serviceguard provides high availability clustering and metro clustering delivering redundancy and rapid recovery for SAP and Oracle environments.
- Enhanced Mission Critical services are available for consolidation, business continuity, modernization and shared services Private Cloud projects. This includes the services to design, deploy and maintain your toughest mission critical environments for maximum resiliency.
I was actually surprised how some of the modular IP innovation was of interest - having common components from power supplies, fans, Virtual Connect and enclosures to new features of HP UX 11i v3. They seemed genuinely impressed with this Mission Critical version of the HP Converged Infrastructure. "Speed & Agility" was important to them, especially for their business critical applications, but one person said it best: "Run fast with confidence". Maybe we should adopt that as our tagline.
You can follow the CI conversation on twitter (#HPCI): www.twitter.com/HPConverge
Hopefully by now you will have started to pick up some basic understanding about what HP is trying to drive around it's new Converged Infrastructure strategy. You will have read that IT sprawl is taking business performance to the breaking point and that 70% of most IT budgets are being spent on operations, and that a shrinking 30% is spent on innovation. This was neatly summarized recently by David Hughes, SAS Vice President International Sales, who stated, “it is going to be increasing more important to be able to leverage complexity as opposed to becoming a victim of it, in a cost constrained environment. This is a growing trend for SAS customers.” (Reference: Technology at Work press day, in Frankfurt Germany on April 26th, 2010).
HP Converged Infrastructure strategy attacks IT sprawl head on, as it takes an holistic approach to datacenter issues, enabling business to tackle the problems of application, server, storage and network which currently inhibits today’s business agility. For more information check out www.hp.com/go/ci
This is a great approach from HP, but what would a pessimistic view look like from an industry analyst perspective. Would the HP strategy hold up against what industry analysts are predicting? I had the opportunity to hear this POV from a leading industry analyst – Andy Butler from Gartner, who presented at the Technology at Work conference in Frankfurt Germany.
Andy summarized that the current datacenter – compute, storage and management are all seen as separate stacks by design. This has created multiple islands, all gated by physical clusters, each having very different limitations. This in turn as has led to slower, more siloed and costly deployments for IT to manage.
Andy defines change happening in the follow key areas:
- Virtualization – The engine room for infrastructure transformation
- Management Tools – The emerging battle ground for control and collaboration
- Infrastructure Convergence – built upon a fabrics based architecture
This fabrics based architecture will comprise of a set of compute, storage, memory and I/O components joined through a common Fabric interconnect.
A Fabrics based architecture will enable you to:
- Only buy what you need when you need them
- Wire once and reconfigure many times
- The capability to manage from a more holistic standpoint across the fabric
A prediction into the future
Andy predicted that by 2012, 30% of global 2,000 datacenters will be equipped with some fabric based bladed architecture
- Corollary 1. 90% of these Fabric systems will employ virtualization
- Corollary 2. 15% of these Fabric systems will be utilized for cloud infrastructures and Services (a gradual trend that will move into the datacenter over the next 5 – 10 years)
- Start now and assess the number of Server platforms you currently have within your datacenters
- Know what you have so you can start to plan around and prepare for a future Fabrics based computing environment
- Multiple influences are driving the server market towards an x86 architecture, with smaller forma factors and modular fabrics