Data Storage Problems

This week the New York Times published a long article about the problem of data storage, and I would like to summarize some of their findings. The article is available here in Saturday’s technology section.

The article is an attack at what the author sees as wasteful use of resources in data storage centres. There are now hundreds of thousands of these huge centres spread throughout the world, and the problem is they use an incredible amount of electricity. The servers have to be kept cool and they have to have spare capacity so that we can download whatever we want whenever we want.

Inside a US data centre

Inside a US data centre

Worldwide these centres use about 30 billion watts of electricity, and that is about 30 nuclear power plants worth of power. A single data center uses about the same amount as a small town, and the main criticism is the nature of the usage.

In the US 2% of all electricity used goes to these data centers, but the vast majority of this resource is wasted. Typically many servers are left to run 24 a day but never or rarely used (more than half in this study), and the average machine in operation uses less than 10% of its capacity. Servers are left running obsolete programs or in ‘comatose’ because nobody wants to risk a mistake and turn them off.

All of this means that any data center might use 30 times as much electricity as is needed to carry out the functions it performs.

All of these centres also have to have a back up in case of power failure, and so are surrounded by diesel generators and stacks of batteries, and many have been found in breach of environmental regulations and fined. The article gives details but the companies are names that we all know and use.

If you read the more than 300 comments however you will discover that a lot of people do not agree with the findings as reported. Many technicians argue that the companies cited are investing huge amounts of money into making the storage of data more efficient, and are constructing wind farms and using solar power in an attempt to cut costs and emissions. The article has its agenda and exploits it fully, but the problem is real.

I personally believe that we are witnessing the results of a digital culture change. We no longer have to store data on our machines, we can store it in some mythical cloud out there in the cyber-universe. This makes us think that it somehow exists without the need for a hard drive, but this is not true. As a result we keep things that we do not need. I have 500 e mails in my inbox, with attachments, photos that I will never again look at and other useless things, and they are all in storage somewhere.

Technology advances, storage gets cheaper and uses less space, but the amount of data created is growing at an incredible rate. My question is, can we do anything about it? Are we not the ones who should take some responsibility and think about the consequences of our actions. We think about not using paper to print emails but we don’t think about not sending them!

Smarter Cities

Following on from my post last week about Apps and Christopher’s post about smart skies on Tuesday of this week I would like to introduce smart cities. Harvard University graduate School of Design run a course that they describe as Urban Cybernetics, called in fact Smart Cities. It is taught by Nashid Nabian, and the aim is to design urban projects that use technology to improve urban life.

The final projects are available to browse online through the course website, and many are very interesting. Cameras and sensors are some of the tools of the trade, used to measure pedestrian or cyclist use of the city, to improve traffic flow or better understand the mechanics of the city from a host of other points of view.

One of my favourite projects is called The Listening City and was written by Carolina Soto within the Real-time Cities course in autumn 2011 as part of the Responsive Environments and Artifacts Lab. Her project involves the use of QR codes that can be scanned using a smartphone. The codes are attached to street furniture, dustbins and all types of functional objects for the running of the city. If a passer by wants to report a problem with any of the objects, they just have to scan it with their smartphone.

 

An example of a QR code in use

A QR code on a traffic signal in Cambridge MA

Once scanned, the user is lead to a website where they complete a questionnaire related to the functionality of the object (in the case of a dustbin if it is full or empty etc, or for a pedestrian crossing light if it works or not).

As the position of each code is known the problem can then be signaled on an interactive map, with the data centrally collected so that it can be analyzed and patterns discovered that may help in improving services.

Take a further look and watch her slide presentation here.

Design and designers affect almost everything in our lives, and working towards improving city life through technology seems like a great idea to me. We are talking about real life improvement that can be seen and measured, information about which bins fill and overflow first is important, the city can be redesigned to confront the problems caused. It might also create solidarity within the residents, more participation in civic life and resolve some of the niggly issues that urban living throws up.

How much data can the human brain hold?

Last week I was listening to a really interesting radio programme, in which I heard a few facts that amazed me. Firstly, did you know that your brain has around one hundred billion (yes, 100,000,000,000) neurons. Each of those neurons are thought have tens of thousands of connections to other neurons. That means in your brain there are… a lot of connections!

Anyhow, in the radio show they also stated that a cubic millimetre of brain tissue contained a petabyte of data. Unless you know your bits and bytes, you might not realise what that means. To put it simply, one cubic millimetre of brain matter has the capacity to store all the digital images currently on Facebook, i.e. every image the social network has ever had uploaded. This is only possible thanks to the number of connections between brain cells.

All that in just one cubic millimetre of your brain! Gosh.

This got me thinking, how much data can the human brain hold? To start investigating this I first needed to brush up on my knowledge of storage units.

Bits And Bytes

Okay, lets start from the beginning. The smallest packet of data you can get is a bit, which is equivalent to one binary digit. A byte is equivalent to 8 bits of data, therefore a bit is equal to 0.125 bytes. Make sense so far?

In between bit an byte is the less well known nibble, which is equivalent to 4 bits. I suspect whoever invented bits, nibbles and bytes was either a little obsessed with food, or quite peckish at the time!

The next step up from a byte is a kilobyte, which is equal to 1,000 bytes in terms of storage space. If you wanted to talk about processing ability, one kilobyte is equal to 1,024 bytes – lets stick to storage though!

1kB = 1,000 (one thousand) bytes

This is still really tiny. To put it into perspective the size of a typical 80 word plain text email is around 10kb (or 10,000 bytes) – source About.com.

From kilobytes we climb to megabytes which equal 1,000 kilobytes.

1MB = 1,000,000 (one million) bytes

On my digital camera, I have it set to the highest quality and image size to 3264 by 2448 which produces images usually between 2 and 5MB. Quite big relative to a kilobyte.

The megabyte has nothing on the gigabyte though, which is 1,000 megabytes!

1GB = 1,000,000,000 (one billion) bytes

The maximum any CD ROM can hold is 900 megabytes of data which is 100MB less than a gigabyte. So a gigabyte is just bigger than a CD.

After a gigabyte comes a 1,000 times bigger terabyte.

1TB = 1,000,000,000,000 (one trillion) bytes

In my current computer I have a terabyte sized hard disk. Many computers come with terabyte or half terabyte hard disks nowadays, however go back seven or eight years and 80gb (8% of a terabyte) was around the average hard disk size, showing the advancements that have been made in just a few years.

Eventually we reach the petabyte, and you guessed it, it is 1,000 times bigger again.

1PB = 1,000,000,000,000,000 (one quadrillion) bytes

One petabyte is the amount of data one cubic millimetre of brain tissue can hold. That’s two petanibbles and eight petabits.

How Much Data Could A Brain Be?

There are so many factors which affect brain size that it is going to be hard for me to work this out with any sort of accuracy, but I will try. Ethnicity, gender and body size along with many other factors affect brain size. The amount of data the brain can store isn’t solely dependant on size, but lets ignore that for now. The average female brain is around 1130 cubic centimetres, whilst the average male brain is 1260 cubic centimetres. That said women have more connections between the two hemispheres than men do.

Averaging it out, that means that the average brain size is 1195 cm3. So how many cubic millimetres go into 1195 cm3? 1,195,000 – to be precise.

The brain is 1.2 zettabytes in sizeThat means that the human brain can store 1,195,000 petabytes of data! That is equivalent to 1.195 zettabytes, as an zettabyte is equal to 1,000,000 petabytes.

What Percentage Of The Brain Is Storage?

Not all of your brain is dedicated to storage though, meaning that 1.195 zettabytes isn’t true to the amount of data we can store. So, how much of the human brain is storage?

From my understanding, most memory processes and storage happens in the temporal lobe. This is approximately 25% of the brain – that is a very approximate percentage!

How Much Data Can A Brain Hold?

In order to answer the above question, I am assuming that 25% of the brain is dedicated to memory, whilst the rest is required for other functioning, and that the average brain is 1,195,000 cubic millimetres in size.

25% of 1,195,000 cubic millimetres is 298,750 cubic millimetres. If one cubic millimetre of brain tissue contains a petabyte of data then the average human brain is able to hold 298,750 petabytes of data.

298,750 petabytes of data is equivalent to:

  • 2,390,000,000,000,000,000,000 bits
  • 298,750,000,000,000,000,000 bytes
  • 298.750 exabytes (approx 300 exabytes)
  • 0.29875 zettabytes (approx 0.3 zettabytes)

So there you have it, were you to max out your memory, you could probably store around 300 exabytes of data.

To put 300 exabytes into perspective, one standard single layer Blu-Ray disc can hold 25GB of data – approximately 5 hours of HD video. Therefore your brains memory could theoretically hold 12,000,000,000 Blu-Ray discs – around 60 billion hours of HD video!

Will we ever be able to technically match the amazing feats that nature has achieved? I am not so sure you know…