IT Sustainability Think Tank: Solving the IT energy efficiency and usage crisis | Computer Weekly

Date:


The current global energy crisis, triggered by a mix of energy supply and transmission system constraints and generator maintenance challenges, has challenged digital infrastructure operators with energy price increases, increased electricity carbon emissions factors, and reduced energy security.

As external events gyrate energy markets, governments and energy suppliers are taking an “all hands-on deck” strategy, establishing price controls and extending or restarting nuclear and fossil fuel generation capacity to keep the lights on. With very little ability to influence the energy markets, what can operators do to manage these cost and reliability challenges? How can operators improve operational environmental performance and sustainability in an environment of increased electricity emissions factors?

Looking beyond the immediate solution of deploying backup generation capacity, which facilitates energy reliability but not cost management or environmental performance improvements, operators have an untapped energy source hidden in plain sight: IT system efficiency improvements.

Digital infrastructure operators can meet energy reduction commitments and reduce their energy-based carbon emissions, even with increases in the grid carbon emissions factors.

With energy cost and supply concerns and increased use of coal projected to continue through 2023 and into 2024, investments in power management deployments, server and storage equipment consolidation, IT equipment refreshes, and automated IT space temperature and humidity control offer near-term paybacks and long-term benefits to the reliability, competitiveness and environmental performance of digital infrastructure operations.

Power management deployment

Anecdotal evidence suggests that fewer than 20% of installed servers operate with power management functions enabled. While this is the right choice for applications with high-performance and low latency requirements, many application types can successfully co-exist with power managed IT infrastructure. Digital infrastructure operators can capture energy use and cost reductions of up to 20%.

IT equipment manufacturers offer power management templates tuned for specific workloads or can assist in designing a specialty template. Operators should test applications to validate their compatibility with these functions and ensure they meet reliability and latency requirements. A large banking company reported that their validation program resulted in the deployment of power management functions on a significant percentage of their server fleet, achieving a 10% reduction in energy use and cost.

Server and storage equipment consolidation

Uptime Intelligence’s annual survey of and conversations with digital infrastructure managers indicate that a significant percentage of the installed server base operates at average utilisations of less than 30%.

Studies and real-world experience demonstrate that servers supporting enterprise and office software can operate at 50-60% utilisation through the adoption of virtualisation technologies.

These technologies are proven and their deployment, facilitated by workload placement software, can improve IT system reliability and resiliency while reducing the hardware and energy footprint by up to 50%.

Savings can be realised by virtualising and consolidating operations on existing IT infrastructure, migrating virtual machines (VMs) to a private or public cloud, or consolidating VMs onto new, higher-capacity IT equipment. While virtualising or further consolidating an IT ecosystem requires effort and expense, the work can produce significant resiliency improvements and cost savings. 

Operators can use capacity optimisation methods (COMs) and tiered storage hierarchies to reduce the storage system footprint and increase the GB of data stored per watt of power consumption. COMs, including data compression, thin provisioning and data de-duplication technologies, increase data density while enabling storage capacity utilisation of 80% or more. Like server consolidation, this can decrease installed equipment footprints and energy use by up to 50%.

Tiered storage hierarchies reduce overall energy use by matching required data availability to storage device response times, with high availability data stored on solid-state storage (SSD) and 15,000 rpm hard disk drive (HDD) with high energy demand, less important data stored on lower energy 10,000 and 7,200 rpm HDDs and archived data stored on energy sipping tape drives. In aggregate, the system meets data availability requirements while minimizing energy use.

Equipment refresh

New IT equipment typically offers higher work or storage capacity and improved work per watt or TB per watt performance characteristics. However, a refresh by itself does not create a more sustainable IT infrastructure. Applications and storage data from two to four existing products must be consolidated onto a single new product. It is possible to attain even higher refresh ratios in situations where utilisation levels on current equipment are low. Consolidation on refresh can deliver savings in hardware footprint and energy use of 50% or more.

Automated IT space temperature and humidity control: IT space temperature can be controlled and maximised by installing temperature and humidity sensors and deploying a software-driven optimisation package. These systems improve IT system reliability by eliminating hot spots and reducing cooling energy use by up to 20% by raising the average datacentre temperature while matching cooling delivery to heat generation.

Full deployment can take several months, as operators must make temperature increases in a stepwise fashion to enable rebalancing of floor tiles to eliminate hot and cold spots within the space.

An IT manager has a menu of energy efficiency opportunities at their fingertips. Capturing savings while maintaining operational reliability and improving environmental performance requires a well-developed project plan implemented by a dedicated, knowledgeable team.

Involvement of and communications across facility, IT, procurement and supplier teams will be critical to the execution and success of the project due to the high level of interdependency between these groups. Even in an IT consolidation project, facilities involvement is critical to rebalance and optimise the operation of the central and IT space cooling systems to capture savings in cooling delivery.

While these projects require significant investment in time and resources, they offer immediate benefits in decreased operational cost, improved environmental performance, and reduced energy demand on a strained energy supply system. The reduction in energy consumption and associated emissions more than compensate for the increase in emissions driven by the temporary rise in fossil fuel generation capacity.

More importantly, the projects establish the operational processes, procedures and mindset that continually maximises the work delivered per watt consumed, improving the environmental performance and sustainability of the digital infrastructure system.



Source link

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Share post:

Popular

More like this
Related

Nvidia: Chip giant posts record sales as boss sees AI ‘tipping point’

The chip giant's shares have soared by more...

How AI is helping the search for extraterrestrial life

Artificial intelligence software is being used to look...

Gemma, Google’s new open-source AI model, could make your next chatbot safer and more responsible

Google has unveiled Gemma, an open-source AI model...