This week, we delve into the world of power distribution units. We also take a look at the correlation between climate change and data centers; we also examine how machine learning can both help and hinder cooling efforts and more news.
Data Center News Roundup for Friday, March 1, 2019
We know, power strips aren’t exactly the sexiest components in the data center, but when chosen poorly, they can cost you a bundle in downtime. There’s no good reason to choose poorly; quality abounds in this niche market. The only question is, how much do you need in the way of bells and whistles on your PDUs? This week, Data Center Frontier outlines the most important factors when choosing power strips, including: form factor, input plugs that match the rack, matching outlets and, of course, functionality. Some data center managers need real-time information on power, which can be handled by smart PDUs, while others only require basic strips. Read the story to determine which is best for your next build or upgrade, then visit Instor for sales and installation of PDUs.
Cloud Cost Management Woes
According to Data Center Knowledge, a new survey shows that while cloud adoption is soaring higher than ever before, most companies aren’t optimizing the service and are missing out on the savings promised by the cloud. The report says that less than half of all AWS customers tap into such cost-saving features as Reserved Instances, with similar figures for those using Microsoft Azure and Google Cloud. Based on a survey of nearly 800 organizations across industries, the findings are collected in the RightScale 2019 State of the Cloud Report. Learn more here.
Data Centers Good for the Earth?
When the topic of data centers and climate change comes up, it’s usually a negative correlation, with the energy-sucking nature of these facilities being partly blames for climate change. However, a piece in Clean Technica looks at some of the ways data centers could work to lower their environmental impacts such as reusing waste heat, tapping into renewable energy and adopting better best practices for energy efficiency. The article even envisions a day when some governments require data centers to run on 100 percent clean energy by phasing out natural gas. Read the breakdown and possible solutions in “Data Centers as a Solution to Climate Change.”
Machine Learning Pros and Cons
These days, machine learning in the data center is something of a double-edged sword. On the one hand, this technology can automate processes and detect problems, even predicting issues before they arise. On the other, this requires a lot of cooling power. And, with the demand for AI and machine learning on the rise, Data Center Frontier writes, the next generation of data centers will have to grapple with balancing the need for speed without breaking the bank on power bills. There are some promising ideas being fleshed out as we speak, including a Google algorithm that will help tame cooling costs along with PID loops that correct cooling in real-time. To peer into the future, check out “Data Center Cooling: Machine Learning is the Problem and the Solution.”
Ready to see how Instor can help with your next fitup or buildout? Click below to get a no-cost estimate.