Data Centre Hardware

When it comes to data centre efficiency and measuring the performance of IT hardware, the Server Efficiency Rating Tool (SERT) 2.0 is the gold standard. The SERT 2.0 Suite delivers a benchmarking tool to assess the efficiency of OEM hardware.  

However, when it comes to OCP equipment, the SERT 2.0 Suite is currently incompatible.  

Techbuyer’s research team have years of experience exploring the efficiency of OEM hardware and this April they’re attending the OCP conference, taking place in Lisbon, to demonstrate how we can work to apply SERT 2.0 to OCP hardware.  

Levelling the Playing Field: Industry Standard Benchmarks and OCP  

With the data centre market capturing the attention of legislative bodies around the world, in large for their environmental impact, but also as a result of the growing demand for data, it is essential for data centres to assess their overall efficiency.  

Various directives, such as EnergyStar in the US and the Energy Efficiency Directive in the EU, mandate reporting on server efficiency using the SPEC SERT 2.0 benchmark.  

This benchmark is used to measure and evaluate the performance and efficiency of servers across different workloads. However, it must be run according to specific configurations, setups and sources that do not comply with OCP technology.  

This means that the OCP community faces challenges in meeting standardised energy efficiency metrics when compared to OEMs. However, Techbuyer’s

Research Team, in partnership with RISE, have proposed a research project to establish SERT 2.0 benchmarking for OCP technology.  

RISE, who have years of experience working with OCP technology, can add to our experience working with and analysing SERT and SPECpower to create a sound and viable methodology for benchmarking OCP technology.  

Using bespoke wind-tunnel technology we can explore the relationship between environment, energy efficiency and performance for server hardware. Whilst SERT 2.0 prohibits the placing of hardware between the power analyser and the server (such as a powershelf), the methodology does support the benchmarking of blade enclosures. As a result, our team are proposing that we build on the similarities to create a testing parity.  

What Could This Mean for the OCP Market?  

Being able to benchmark efficiency according to accepted markets allows the OCP market to match OEM hardware and remain competitive. As legislative pressures on the efficiency of technology increase, measurement techniques must be developed and adapted. This project offers a gateway into this process, as well as a potential solution into OCP energy usage assessment.  

What Does This Mean for Our Customers? 

We bring our research directly into the services you receive, including configure-to-order server builds and technical advice. This means that you receive expertise for every IT decision you need to make, paired with industry-leading research findings. Find out about our full range of services and how we can help you create a bespoke IT solution here.