See how TSO Logic
simplifies IT transformation.

Our software platforms remove the guesswork so you can now plan with confidence.

reCAPTCHA is required.

See how TSO Logic
simplifies IT transformation.

Our software platforms remove the guesswork so you can now plan with confidence.

reCAPTCHA is required.

Shift the paradigm. It’s not all about PUE.

Posted in: Blog Industry Insight

Power Usage Effectiveness (PUE) is a facility-focused metric that measures cooling of the heat generated by servers, storage devices, and other equipment. While it doesn’t present a holistic view of efficiency levels, PUE has, for the past decade or so, been the industry-standard metric used — even overused — to measure data center efficiency. But the time has come to shift the paradigm to a broader view.

Data center expenses are often a company’s single largest expense. So even executives who know little about its internal workings are inclined to lay down pressure to reduce costs. With actionable intelligence you can make improvements that reduce heat, and costs, in your data center — and transform an energy-hungry cost center into a strategic business asset.

Move beyond one-and-done measurements on the facility side

With a more robust view inside the data center, you’ll get a better understanding of IT workloads and be able to locate the source of energy waste. Our agentless software discovers application-level information, showing which applications are generating revenue and supporting the business and where you might be wasting money on idle servers and licensing fees. Remember, you can’t manage what you can’t measure.

Simple solution, big results

Moving beyond traditional PUE enables companies to make significant improvements and realize major savings. With actionable intelligence you gain the perspective needed to better provision and use automation or orchestration to cut back on underutilized or forgotten equipment. Here are just a few cases we’ve encountered:

Server capacity: active versus idle
Idle capacity is higher than you think. In the 15 non-HPC data centers we recently measured, idle capacity was 45 to 55%. By measuring application and server usage on an ongoing basis and using automation to drive down excess capacity, these data centers are eliminating waste, improving efficiencies, and meeting their SLAs..

Abandoned and forgotten servers
They’re out there, but not necessarily easy to find. We’ve worked in environments where 20% or more are powered on but not doing any useful work — such as host servers with no guests deployed.

Application provisioning
We’ve seen it consistently: Older, less efficient servers are utilized more than newer, more efficient ones. With the right measurements and automation you can drive applications and workloads to your most efficient servers — and the savings are BIG.

Measure the implications of your storage policies, including how storage is accessed and what types of data are stored on which platforms. Check to see if rarely accessed data is stored on fast spinning disks that are expensive to run. An updated infrastructure can pay for itself in short order.

Moving beyond PUE can help you achieve a balanced and efficient data center that scales with your business and supports your SLAs. Amazon, Google, and Microsoft figured this out long ago, but have kept their innovations proprietary. Today, you can easily reach or exceed their levels of cost and power efficiency without having to build it yourself. The tools are readily available. Just ask.

Related post: What you can learn from cloud providers. Read more.


Aaron Rallo
CEO and Founder of TSO Logic