DZone
Thanks for visiting DZone today,
Edit Profile
  • Manage Email Subscriptions
  • How to Post to DZone
  • Article Submission Guidelines
Sign Out View Profile
  • Post an Article
  • Manage My Drafts
Over 2 million developers have joined DZone.
Log In / Join
Please enter at least three characters to search
Refcards Trend Reports
Events Video Library
Refcards
Trend Reports

Events

View Events Video Library

Zones

Culture and Methodologies Agile Career Development Methodologies Team Management
Data Engineering AI/ML Big Data Data Databases IoT
Software Design and Architecture Cloud Architecture Containers Integration Microservices Performance Security
Coding Frameworks Java JavaScript Languages Tools
Testing, Deployment, and Maintenance Deployment DevOps and CI/CD Maintenance Monitoring and Observability Testing, Tools, and Frameworks
Culture and Methodologies
Agile Career Development Methodologies Team Management
Data Engineering
AI/ML Big Data Data Databases IoT
Software Design and Architecture
Cloud Architecture Containers Integration Microservices Performance Security
Coding
Frameworks Java JavaScript Languages Tools
Testing, Deployment, and Maintenance
Deployment DevOps and CI/CD Maintenance Monitoring and Observability Testing, Tools, and Frameworks

Because the DevOps movement has redefined engineering responsibilities, SREs now have to become stewards of observability strategy.

Apache Cassandra combines the benefits of major NoSQL databases to support data management needs not covered by traditional RDBMS vendors.

The software you build is only as secure as the code that powers it. Learn how malicious code creeps into your software supply chain.

Generative AI has transformed nearly every industry. How can you leverage GenAI to improve your productivity and efficiency?

Related

  • How to Accelerate Hyper-Automation With Industrial IoT
  • 4 Best Practices for IoT OTA Updates
  • Top 10 IoT Development Tools in 2020
  • Next-Gen IoT Performance Depends on Advanced Power Management ICs

Trending

  • ITBench, Part 1: Next-Gen Benchmarking for IT Automation Evaluation
  • Apache Spark 4.0: Transforming Big Data Analytics to the Next Level
  • Caching 101: Theory, Algorithms, Tools, and Best Practices
  • How to Format Articles for DZone
  1. DZone
  2. Data Engineering
  3. Data
  4. Mastering PUE for Unmatched Data Center Performance

Mastering PUE for Unmatched Data Center Performance

PUE can make or break data center energy efficiency. Improving this figure will help your facility lower costs and become more sustainable.

By 
Emily Newton user avatar
Emily Newton
·
Aug. 20, 24 · Analysis
Likes (1)
Comment
Save
Tweet
Share
4.1K Views

Join the DZone community and get the full member experience.

Join For Free

Data centers use a lot of electricity. Consequently, as they grow, it becomes all the more important to understand their energy use and take steps to reduce it. Learning to measure data center power usage effectiveness (PUE) is a crucial part of that goal.

What Is Data Center Power Usage Effectiveness?

As its name implies, PUE measures how efficiently a data center uses the energy it consumes. It represents this figure as a ratio between the facility’s total electrical usage and how much of it goes toward IT equipment. The more even the ratio, the better.

The term first emerged when The Green Grid was founded in 2007, bringing attention to the sustainability of the burgeoning data center industry. Since then, cloud adoption has only grown, making data center PUE more relevant than ever. However, some argue the metric is losing reliability as a measure of efficiency as better PUEs become more common.

The argument against PUE centers around its limited scope. While an eco-friendly data center will divert most — if not all — of its power to IT hardware, that doesn’t mean all IT power usage is efficient. PUE does not show how the IT processes themselves could become greener, so it may be misleading. The metric also often relies on estimates, leaving room for error.

Despite these shortcomings, PUE is still a helpful tool. It may not be a complete measure of your facility’s efficiency, but it’s a good start.

How To Calculate Data Center PUE

The data center PUE calculation is fairly straightforward. You take your total consumption and divide it by that of your IT infrastructure. However, doing so requires measuring the electricity usage of everything in your facility — including your hardware, lighting, electricity distribution systems, PSUs, cooling, and miscellaneous losses.

Datacenter PUEs today hover around 1.8 on average, suggesting IT equipment often accounts for just over half of total energy consumption. Such a gap leaves considerable room for improvement, especially considering how some facilities can achieve a 1.06 PUE or lower.

Why Does PUE Matter?

While data center power usage effectiveness may be an imperfect measure, it remains crucial for businesses today. As AI and other digital technologies skyrocket, the industry’s environmental impact is an increasingly worrying concern. Data centers consume 3% of the world’s energy, producing substantial greenhouse gas emissions.

In the long run, facilities must transition to renewables. However, they must address their waste before they do, as renewables are intermittent, and switching will be easier with lower overall consumption. Measuring your PUE is a critical first step in reducing that waste, enabling a greener future.

Tips To Improve Data Center PUE

Once you’ve performed a data center PUE calculation, you should seek to lower the figure. The most impactful changes will depend on which of your processes are the most power-hungry, but there are a few common culprits to address.

1. Embrace Passive Cooling Methods

Cooling is often the largest issue, as HVAC accounts for 25% to 40% of all electricity in data centers. At the same time, you don’t want to sacrifice temperature control, as high temps will lead to downtime. The solution is to implement a few passive cooling methods.

Start by placing racks so the backs of alternating rows face each other. This orientation will create hot aisles where each server’s exhaust meets and cold ones where the fronts are. Separating hot and cool air like this enables easier airflow, reducing HVAC strain and related consumption.

Raising servers off the floor will further improve airflow. Cable management offers another place to embrace passive cooling, as minimizing clutter will prevent overheating.

2. Address Power Distribution

Your energy distribution deserves attention, too. Losses in these systems account for 12% of total consumption, so a more efficient power distribution unit (PDU) will make a big difference.

Automation is the key here. Smart PDUs can automatically adjust power cycling and resource allocation according to real-time conditions and demands. As a result, they use as little electricity as possible while maintaining operations. Similarly, automated load balancing and virtualization can minimize strain to prevent losses.

Maintenance is another crucial consideration. Using Internet of Things sensors to catch repair issues as they emerge will enable faster fixes to keep PDUs in prime condition for longer. Consequently, they’ll experience less waste.

3. Upgrade Your UPS

Next, review your servers’ uninterruptible power supply (UPS). Many data centers rely on older, outdated models, as they can last for several years before needing replacement. However, just because a UPS is still functional does not mean it’s efficient.

Upgrading to a newer UPS can eliminate many losses, ensuring more of your total consumption goes toward IT functions. Some modern UPSes do not feature any transformers, minimizing the number of components in the circuit — fewer obstacles to pass through lead to less waste.

Fast cycle times and small filters are other UPS features to look for. Ideally, your UPS should also produce minimal heat, as this will lower your cooling requirements and reduce HVAC energy expenditure.

4. Revisit Your Lighting

Some aspects of data center PUE are less obvious. Lighting, for example, is typically not a major source of electricity withdrawals, but it is often wasteful. However, many organizations miss this optimization opportunity because it does not appear significant at first.

Switching to LED lights is a crucial first step. LEDs are up to 90% more efficient than incandescent alternatives, thanks largely to their minimal heat production. While such improvements may seem marginal in comparison to HVAC figures, every step forward represents lower costs and increased sustainability.

It’s also worth considering that data centers only need light when employees are in the server room. Consequently, it’s best to use a motion sensor to turn lights on when someone enters and off when people leave.

5. Monitor PUE Over Time

After making any adjustments, remember to perform another data center PUE calculation. The only way to ensure your changes are effective is to monitor the same key performance indicators (KPIs) before and after.

Take note of which strategies yield the biggest PUE improvements. Identifying these leaders will make it easier to perform similar upgrades to future data centers. Similarly, record any adjustments that do not produce meaningful results. You may save time and money by avoiding them in later projects.

Keep monitoring the same KPIs as your data center grows. Additional servers, new hardware, or higher workloads may change your PUE, requiring further optimization. Recognizing the opportunity to improve earlier will help you maximize performance.

Maximize Your Data Center PUE Today

Your data center power usage effectiveness impacts your operational expenses and sustainability. Both are crucial areas to address, so PUE demands your attention. Learning how to use this ratio and where you can improve it is the first step to forming a more efficient and profitable facility.

IT Data (computing) Performance Environmentally friendly IoT

Opinions expressed by DZone contributors are their own.

Related

  • How to Accelerate Hyper-Automation With Industrial IoT
  • 4 Best Practices for IoT OTA Updates
  • Top 10 IoT Development Tools in 2020
  • Next-Gen IoT Performance Depends on Advanced Power Management ICs

Partner Resources

×

Comments
Oops! Something Went Wrong

The likes didn't load as expected. Please refresh the page and try again.

ABOUT US

  • About DZone
  • Support and feedback
  • Community research
  • Sitemap

ADVERTISE

  • Advertise with DZone

CONTRIBUTE ON DZONE

  • Article Submission Guidelines
  • Become a Contributor
  • Core Program
  • Visit the Writers' Zone

LEGAL

  • Terms of Service
  • Privacy Policy

CONTACT US

  • 3343 Perimeter Hill Drive
  • Suite 100
  • Nashville, TN 37211
  • support@dzone.com

Let's be friends:

Likes
There are no likes...yet! 👀
Be the first to like this post!
It looks like you're not logged in.
Sign in to see who liked this post!