Skip to main content

Optimizing Data Centers Through Machine Learning

Google has published a paper outlining their approach on using machine learning, a neural network to be specific, to reduce energy consumption in their data centers. Joe Kava, VP, Data Centers at Google also has a blog post explaining the backfround and their approach. Google has one of the best data center designs in the industry and takes their PUE (power usage effectiveness) numbers quite seriously. I blogged about Google's approach to optimize PUE almost five years back! Google has come a long way and I hope they continue to publish such valuable information in public domain.



There are a couple of key takeaways.

In his presentation at Data Centers Europe 2014 Joe said:  
As for hardware, the machine learning doesn’t require unusual computing horsepower, according to Kava, who says it runs on a single server and could even work on a high-end desktop.
This is a great example of a small data Big Data problem. This neural network is a supervised learning approach where you create a model with certain attributes to assess and fine tune the collective impact of these attributes to achieve a desired outcome. Unlike an expert system which emphasizes an upfront logic-driven approach neural networks continuously learn from underlying data and are tested for their predicted outcome. The outcome has no dependency on how large your data set is as long as it is large enough to include relevant data points with a good history. The "Big" part of Big Data misleads people in believing they need a fairly large data set to get started. This optimization debunks that myth.

The other fascinating part about Google's approach is not only they are using machine learning to optimize PUE of current data centers but they are also planning to use it to effectively design future data centers.

Like many other physical systems there are certain attributes that you have operational control over and can be changed fairly easily such as cooling systems, server load etc. but there are quite a few attributes that you only have control over during design phase such as physical layout of the data center, climate zone etc. If you decide to build a data center in Oregon you can't simply move it to Colorado. These neural networks can significantly help make those upfront irreversible decisions that are not tunable later on.

One of the challenges with neural networks or for that matter many other supervised learning methods is that it takes too much time and precision to perfect (train) the model. Joe describing it as a "nothing more than series of differential calculus equations " is downplaying the model. Neural networks are useful when you know what you are looking for - in this case to lower the PUE. In many cases you don't even know what you are looking for.

Google mentions identifying 19 attributes that have some impact on PUE. I wonder how they short listed these attributes. In my experience unsupervised machine learning is a good place to short list attributes and then move on to supervised machine learning to fine tune them. Unsupervised machine learning combined with supervised machine learning can yield even better results, if used correctly.

Comments

Popular posts from this blog

15 YEARS OLD GIRL IMPREGNATED AND MAN RESPONSIBLE FOR IT TOOK FOR AN ABORTION THAT FAILED

BBI FACILITATE ARREST OF 35 YEARS OLD FOR DEFILEMENT, IMPREGNATING 15 YEARS OLD GIRL AND ABORTING FIVE MONTHS PREGNANCY IN ANAMBRA STATE. Today, at 1:26pm, We received a complaint from a concerned citizen who informed us of a 15yrs old girl brought into a hospital for medical treatment. Our intelligence team led by Director General Gwamnishu Emefiena Harrison Kenneth Nwaobi Ezika Kene and others left Asaba and arrived Ogidi Anambra state for investigation. 35yrs Chris Azuoma took the victim to hospital where she was injected and given abortion pills. She bled heavily and had complications and so decided to take her to a specialist hospital to evacuate the foetus. Getting to the hospital, we met the management and identified ourselves as Human rights group and they granted us permission to interview the victim. She confirmed the story and the perpetrator confessed forcefully having unprotected sexual intercourse with the victim. 2015 Administration of Criminal Justice permit private per

Hacking Into The Indian Education System Reveals Score Tampering

Debarghya Das has a fascinating story on how he managed to bypass a silly web security layer to get access to the results of 150,000 ISCE (10th grade) and 65,000 ISC (12th grade) students in India. While lack of security and total ignorance to safeguard sensitive information is an interesting topic what is more fascinating about this episode is the analysis of the results that unearthed score tampering. The school boards changed the scores of the students to give them "grace" points to bump them up to the passing level. The boards also seem to have tampered some other scores but the motive for that tampering remains unclear (at least to me). I would encourage you to read the entire analysis and the comments , but a tl;dr version is: 32, 33 and 34 were visibly absent. This chain of 3 consecutive numbers is the longest chain of absent numbers. Coincidentally, 35 happens to be the pass mark. Here's a complete list of unattained marks - 36, 37, 39, 41, 43, 45, 47, 49, 51, 53,

Reveiw: Celluon Epic Laser Keyboard

The Celluon Epic is a Bluetooth laser keyboard. The compact device projects a QWERTY keyboard onto most flat surfaces. (Glass tabletops being the exception) You can connect the Epic to vertically any device that supports Bluetooth keyboards including devices running iOS , Android , Windows Phone, and Blackberry 10. On the back of the device there is a charging port and pairing button. Once you have the Epic paired with your device it acts the same as any other keyboard. For any keyboard the most important consideration is the typing experience that it provides. The virtual keyboard brightness is adjustable and is easy to see in most lighting conditions. Unfortunately the brightness does not automatically adjust based on ambient light. With each keystroke a beeping sound is played which can be turned down. The typing experience on the Epic is mediocre at best. Inadvertently activating the wrong key can make typing frustrating and tiring. Even if you are a touch typist you'll still