Hey—we've moved. Visit
The Keyword
for all the latest news and stories from Google
Official Blog
Insights from Googlers into our products, technology, and the Google culture
A power plant for the Internet: our newest data center in Alabama
June 24, 2015
Every time you check your Gmail, search on Google for a nearby restaurant, or watch a YouTube video, a server whirs to life in one of our data centers. Data centers are the engines of the Internet, bringing the power of the web to millions of people around the world. And as millions more people come online, our data centers are growing, too.
We’ve recently expanded our data centers in
Iowa
,
Georgia
,
Singapore
and
Belgium
. And today we’re announcing
a new data center
in Alabama—our 14th site globally.
This time, we’re doing something we’ve never done before: we’ll be building on the grounds of the
Widows Creek coal power plant
in Jackson County, which has been scheduled for shutdown. Data centers need a lot of infrastructure to run 24/7, and there’s a lot of potential in redeveloping large industrial sites like former coal power plants. Decades of investment shouldn’t go to waste just because a site has closed; we can repurpose existing electric and other infrastructure to make sure our data centers are reliably serving our users around the world.
At Widows Creek, we can use the plants’ many electric transmission lines to bring in lots of renewable energy to power our new data center. Thanks to an arrangement with Tennessee Valley Authority, our electric utility, we’ll be able to scout new renewable energy projects and work with TVA to bring the power onto their electrical grid. Ultimately, this contributes to our goal of being powered by 100% renewable energy.
In 2010, we were one of the first companies outside of the utility industry to
buy large amounts of renewable energy
. Since then, we’ve become the largest corporate renewable energy purchaser in the world (in fact we’ve bought the equivalent of over
1.5 percent of the installed wind power capacity in the U.S.
). We're glad to see this trend is
catching on
among other companies.
Of course, the cleanest energy is the energy you don’t use. Our Alabama data center will incorporate our state-of-the-art energy efficiency technologies. We’ve built our own
super-efficient servers
, invented
more efficient ways to cool our data centers
, and even used advanced
machine learning
to squeeze more out of every watt of power we consume. Compared to five years ago, we now get 3.5 times the computing power out of the same amount of energy.
Since the 1960s, Widows Creek has generated power for the region—now the site will be used to power Internet services and bring information to people around the world. We expect to begin construction early next year and look forward to bringing a Google data center to Alabama.
Posted by Patrick Gammons, Senior Manager, Data Center Energy and Location Strategy
Better data centers through machine learning
May 28, 2014
It’s no secret that we’re obsessed with saving energy. For over a decade we’ve been
designing and building data centers
that use half the energy of a typical data center, and we’re always looking for ways to reduce our energy use even further. In our pursuit of extreme efficiency, we’ve hit upon a new tool: machine learning. Today we’re releasing a
white paper
(PDF) on how we’re using
neural networks
to optimize data center operations and drive our energy use to new lows.
It all started as a 20 percent project, a Google tradition of carving out time for work that falls outside of one’s official job description. Jim Gao, an engineer on our data center team, is well-acquainted with the operational data we gather daily in the course of running our data centers. We calculate
PUE
, a measure of energy efficiency, every 30 seconds, and we’re constantly tracking things like total IT load (the amount of energy our servers and networking equipment are using at any time), outside air temperature (which affects how our cooling towers work) and the levels at which we set our mechanical and cooling equipment. Being a smart guy—our affectionate nickname for him is “Boy Genius”—Jim realized that we could be doing more with this data. He studied up on machine learning and started building models to predict—and improve—data center performance.
The mechanical plant at our facility in The Dalles, Ore. The data center team is constantly tracking the performance of the heat exchangers and other mechanical equipment pictured here.
What Jim designed works a lot like other examples of machine learning, like speech recognition: a computer analyzes large amounts of data to recognize patterns and “learn” from them. In a dynamic environment like a data center, it can be difficult for humans to see how all of the variables—IT load, outside air temperature, etc.—interact with each other. One thing computers are good at is seeing the underlying story in the data, so Jim took the information we gather in the course of our daily operations and ran it through a model to help make sense of complex interactions that his team—being mere mortals—may not otherwise have noticed.
A simplified version of what the models do: take a bunch of data, find the hidden interactions, then provide recommendations that optimize for energy efficiency.
After some trial and error, Jim’s models are now 99.6 percent accurate in predicting PUE. This means he can use the models to come up with new ways to squeeze more efficiency out of our operations. For example, a couple months ago we had to take some servers offline for a few days—which would normally make that data center less energy efficient. But we were able to use Jim’s models to change our cooling setup temporarily—reducing the impact of the change on our PUE for that time period. Small tweaks like this, on an ongoing basis, add up to significant savings in both energy and money.
The models can predict PUE with 99.6 percent accuracy.
By pushing the boundaries of data center operations, Jim and his team have opened up a new world of opportunities to improve data center performance and reduce energy consumption. He lays out his approach in the white paper, so other data center operators that dabble in machine learning (or who have a resident genius around who wants to figure it out) can give it a try as well.
Posted by Joe Kava, VP, Data Centers
Labels
accessibility
41
acquisition
26
ads
131
Africa
19
Android
58
apps
419
April 1
4
Asia
39
books + book search
48
commerce
12
computing history
7
crisis response
33
culture
12
developers
120
diversity
35
doodles
68
education and research
144
entrepreneurs at Google
14
Europe
46
faster web
16
free expression
61
google.org
73
googleplus
50
googlers and culture
202
green
102
Latin America
18
maps and earth
194
mobile
124
online safety
19
open source
19
photos
39
policy and issues
139
politics
71
privacy
66
recruiting and hiring
32
scholarships
31
search
505
search quality
24
search trends
118
security
36
small business
31
user experience and usability
41
youtube and video
140
Archive
2016
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2015
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2014
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2013
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2012
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2011
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2010
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2009
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2008
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2007
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2006
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2005
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
2004
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Feed
Google
on
Follow @google
Follow
Give us feedback in our
Product Forums
.