Is it time to change the server?

The growing availability of broadband internet access across Australia and New Zealand will have a huge impact on how people live their lives and run their businesses. Network data volumes are set to quadruple by 2016. Such ubiquity of access is set to bring the concept of the Internet of Things (IOT) to reality - with everything interconnected and analysed. From tracking your online purchases to devices that monitor their own life cycle and order their own replacement parts.

Cloud computing addresses the need for flexible access, processing power for social networks, data sharing, and business application mobility. While we enjoy the benefits of cloud, we don’t always realise the consequence of consolidation of IT into massive data centres.

Growth in power densities and cooling requirements is going off the scale. For businesses, the key to staying competitive is to do more with less. This requires ever more efficient IT to drive additional value to the business. IT manager pain points are increasingly stringent SLAs, new application builds, higher levels of availability, despite budget constraints.

Today more than 46 per cent of physical x86 servers are implemented as hosts for virtualised environments. Projections to 2016 see that figure rising to 71 per cent.

With an ever increasing number of applications using an ever-decreasing number of physical servers, IT departments will end up with staff burnout.

But there is light at the end of the tunnel for the IT department. Tomorrow’s systems and solutions will be developed around human needs and activities rather than processes. This will shift infrastructure, applications, plus operation and maintenance, to new levels of optimisation.

To evolve with the needs of business, server design must change from boosting application performance to delivering more efficient IT functionality.. It means high integration of storage, servers and I/O; high reliability and minimum management; plus energy and cooling savings on a scale never seen before. This will optimise growth and ensure IT environments deliver new business advantages more quickly.

Fujitsu, a global leader in cloud computing, and developer of server systems for over 60 years has been using some of it’s over $2 billion annual R&D budget to re-vision the server for this new human-centric world.

In implementing its vision for the future, Fujitsu has focused on four essential areas - total server lifecycle management, converged performance, enhanced energy optimization and lifecycle excellence on a global scale.

Total server lifecycle management

Fully automated IT operation is essential in providing highest return on customer investment, cost removal and the freeing of the IT department from complex yet routine tasks. Reduced management complexity means the use of automated, holistic, lifecycle management - as an essential server component - ending the reliance on general software add-ons. Increased use of solid state technology will also offer higher reliability with longer periods between system failures.

Converged performance

40 per cent improvement in performance per watt used is seen as ideal for immediate improvement of business workloads and return on investment. To confirm such real added-value for customers, competitive testing of system performance must be part of the server build process. Top results in virtualisation, energy-efficiency, java web and ERP performance, using industry standard benchmarks, are the most beneficial and understandable way to validate performance benefits. Further, more native integration of servers and storage will remove bottlenecks and accelerate data-intensive applications as data volumes grow.

Enhanced energy optimisation

Sustainability must be viewed as an economic topic with a goal to halve the power and performance burden in the short term. Complex air-conditioning, ventilation and power backup have become major costs on business.

A New York Times study showed the world’s current population of 33 million servers consumes up to 1.5 per cent of the world’s power - the global emissions Argentina or the Netherlands.

But annual data centre power costs continue to increase by 20 percent globally, with Gartner predicting an increase to 50 per cent in the near-term. Server design has to change to improve power supply and power cable efficiencies, plus use low power components and automated power management. Only such highly efficient infrastructures can halve power and air-conditioning costs as well as enhance computer performance. Plus the resulting cool operation has potential in doubling the lifespan of all components.

Lifecycle excellence

Such increases in performance, efficiency and availability will enable hardware related costs to fall below 17 per cent of all IT costs. Now the main IT costs focus must turn to operating infrastructure and new solution implementation. IT suppliers must supply wide-ranging portfolios of services and tools that, reduce those costs throughout the lifecycle, shorten project times and increase application and service availability.

Reference architectures will not only reduce overall risk, but the cost of actual new solution planning and deployment. Comprehensive cloud based data centre integration and logistics services also have the ability to greatly reduce rollout costs by up to 45 per cent.

Paul Hendery is sales and marketing manager at Fujitsu Australia