Blog

A New Hyperscale Imperative For Sourcing IT Hardware

Changing To A Circular Data Center Model With Sesame by ITRenew

The growth of computing with hyperscale technology has made 30 to 40 years of innovation available to all data center operators. But as natively hyperscale applications and business models proliferate, so do environmental impacts of the exploding data center footprint. In fact, 75% of the lifetime “embodied” energy is consumed in the pre-use phase: sourcing raw materials and manufacturing IT hardware. To maximize sustainability, users of hyperscale technology and other operators need to implement models around the circular economy that maximize the lifetime, and reduces the overall embodied energy, of IT hardware.

IT hardware has evolved from shared multiprocessing, through parallel processing, to the hyperscale model

The evolution to today’s hyperscale data center has put enormous compute power into general use that was unthinkable even a decade ago. But scaling up all this IT hardware has made sustainability a major challenge as it creates operational obstacles and sometimes a barrier to growth.

The most recent achievements in the data center evolution started with large scale systems using shared multiprocessing. The goal was to build large nodes that provided very efficient cost, high reliability, and also very high performance. The generation that followed sought to create extremely parallel applications that ran across a much larger number of commodity computer systems. Instead of having one or two or ten large nodes, data centers employed fleets of 10, 50 or a hundred nodes, which were targeted at very specific applications.

The parallel computing phase evolved into the scale-out generation using hyperscale technology. Over the last 10 years many more applications were moved to this model. That happened thru investments made to rethink architectures and rewrite application stacks. Now, all these applications could be deployed in distributed systems with a thousand nodes, or 5,000 nodes, and even 10,000 node clusters. That represents easily two orders of magnitude increases in IT hardware, including servers, storage, networking, and power distribution resources.

Today, modern applications are being explicitly designed for these environments. New business services are developed and deployed with hyperscale technology in mind. They are hyperscale or cloud native by design. These shifts translate into significant benefits for the business community; widespread access to over 30 years of computing innovation, available at large scale quickly and easily.

Hyperscale technology growth balloons the environmental footprint

But, as natively hyperscale applications and business models proliferate, so does the environmental impact of explosive data center capacity. Today, the global IT industry is the third largest global carbon emitter, responsible for 50 million metric tons of annual global e-waste, and that number continues to grow

While data center operators have made great strides in reducing the energy it takes to run and cool a data center, that’s only the beginning. Out of the total energy impact of building, using, and disposing of IT hardware — its lifetime or embodied energy — 75% is consumed in the pre-use phase. Pre-use phase describes everything that happens before the hardware is ever turned on, including the mining of raw materials, transportation, manufacturing, configuration, and installation. In other words, just building a server consumes 75% of its lifetime energy footprint. Approximately 20% of the lifetime energy is consumed during operation of the IT hardware (the use phase) and, finally, a few percent are expended in the post-use phase.

Given the sheer size and number of data centers worldwide, the data center industry as a whole needs to look for reductions across all of those stages. But clearly, the best opportunity to reduce the environmental impact of hyperscale computing is in the pre-use phase. To maximize sustainability, hyperscale and other operators need to implement a circular model that maximizes lifetime utility and value, while reducing the overall embodied energy consumption of data center hardware globally.

A circular data center model using Sesame by ITRenew can limit pre-use impacts

In a circular data center model, the industry is able to create efficiencies across the entire chain of sourcing, using, and reusing. The circular data center model creates secondary lives, uses and markets for IT hardware components. It enables data center operators to amortize embodied energy costs and value across the longest possible useful life. That means for a given workload, we can reduce the collective environmental costs of embodied energy across the entire pre-use phase.

During the use phase, the circular data center can also operate more efficiently. Using advanced hyperscale computing technology, more input power is converted to useful workload than spent on overhead such as cooling and power conversion. We are able to get the most benefit by dedicating maximum input resources toward useful application work, not overhead.

In the post-use phase, the circular data center model allows for technology to be cascaded and new value to be realized. This includes re-engineering and creating secondary uses for rack-based equipment — through solutions such as Sesame by ITRenew — as well as re-used of valuable components that can deliver significant new utility through a variety of consumer computing and edge computing markets.

The circular data center model with Sesame by ITRenew provides the means to fully optimize value and productivity across this entire chain. In the process, all aspects of Total Cost of Ownership (TCO) are addressed in a comprehensive way. Organizations can reduce acquisition costs and operational costs while addressing the pre-use and post-use phases of their IT infrastructure.

Using the circular model provides a powerful baseline for sustainability. Fortunately for today’s data center operators, there is no cost to moving to this mode of thinking. In fact, for organizations that have selected Sesame by ITRenew, they are able to achieve immediate hard dollar savings by maximizing the value of their IT hardware. They can enjoy the benefits of hyperscale computing and minimize the environmental impacts at the same time.

Watch to learn more about the case for circular data centers that create higher proven lifetime value and lower your TCO more and about Sesame by ITRenew and its TCO-smashing rack-scale solutions.