The Monetary Value of Energy in the Data Center

With the interest of Greening a Data Center, the cost of energy has a behavior impact on your organization.

Lewis Curtis and I were chatting earlier today and he posted his ideas on his blog.  It gets you thinking and to save you time I cut and paste the complete post below.  I am working on my own ideas and it will be out next month.

Datacenter energy consumption is rarely a operating cost issue.

From a chat with Dave O'hara today,  I thought I would blog some thoughts around datacenter energy consumption and some common confusion concerning costs.

Do organizations with dedicated datacenters save money when they install more efficient servers and reduce energy consumption? 

Short Answer:  Rarely

Why?

It all is associated with how most dedicated datacenters negotiate energy consumption with utility companies (operating cost issue).   Usually, they negotiate rates at blocks of energy consumption in fixed buckets.

Therefore rule #1:  do not run out of energy,  rule #2: do not leave energy supply stranded (pay for it and not use it).   In other words:  Overprovisioning and Underprovising

Overprovisioning:  you run out of energy in the datacenter (worst sin in the datacenter)

  • TTM (time to market) is significantly impacted
  • Datacenter systems become brittle  (small energy changes can down center)

Underprovisioning: you strand energy your organization already paid for: (very bad)

  • company is wasting money (that's money that could be going into something useful)

There is fine balancing act to managing these issues in your datacenter.

At the end of the day,  we usually have a fixed bucket of energy consumption at a given rate which manage for our datacenter.

So why architect IT solutions which reduces the consumption of energy? 

Answer: Reducing the velocity of datacenter expansion in your organization (capital costs)

As your organization grows in scale and complexity at a given rate, the need build more and more competitive  IT solutions increases at a related velocity.   

As IT solutions expands, organizations need more datacenter capacity to accommodate the business's growth needs.   This expansion is only successful as long as the given IT solutions provide value above the capital of costs of building and operating new datacenter capacity. 

Some Challenges of building new datacenters:

  • Regulations and oversight cost are increasing
  • Cost of datacenter infrastructure are significantly increasing  (PDUs, Cooling solutions, etc..)
  • Negotiated blocks of energy consumption are significantly increasing in price

Therefore, energy efficiency ultimately is about slowing the velocity of datacenter expansion for organization (capital costs).

If you can slow your build cycle from 1 new datacenter every 9 months to 1 datacenter to every 13 months can be a significant business value for your organization.    Slowing forecasted datacenter expansion velocity.

And ultimately, this is not only good for the environment, it's good for the bottom line. 

Lewis