April 5, 2013 | Written by: Shamim Hossain
Share this post:
Digital data is growing at a staggering rate. Many factors and sources are contributing to this growth. The number of connected devices (Internet of Things), explosion of handheld and mobile devices, social and market trends, Web 2.0, social and collaborative tools, availability and penetration of the Internet are just a few. Only cloud can cope with the data that will be generated at a mind-boggling rate and quantity.
According to “The Internet of Things Infographic” published by Intel, 31 billion devices and four billion people will be connected to the Internet by 2020. This equates to almost eight devices per person. I become overwhelmed when I think about the amount of data generated from these devices alone and processing this data in some data centers, most of which will be cloud based in coming years.
Eight bits form a byte. Then we have kilobyte, megabyte, gigabyte, terabyte and so on. The hard disk drive on my first PC had less than 100 megabyte capacity. Now my laptop is equipped with one terabyte. Terabyte was considered a large amount of data even a few years ago. Today sensors from a Boeing jet engine generate 20 terabytes of data every hour.
After terabyte comes petabyte(1015), exabyte(1018), zettabyte(1021) and yottabyte(1024). How big are these?
In an article entitled “What Is the Memory Capacity of the Human Brain?” Paul Reber mentioned that the human brain has a capacity of around 2.5 petabytes of binary data, enough to store three million hours of TV shows. One has to let the TV run continuously for more than 300 years to use up this storage. I wish I could utilize the full capacity of my brain!
What about exabyte? Data generated on Internet each day is about an exabyte. This is equivalent to data stored in about 250 million DVDs.
IBM is developing data management and analysis technologies for Square Kilometer Array (SKA), the world’s largest and most sensitive radio telescope. SKA is expected to generate more than one exabyte of data everyday. To put this into context, IBM Australia Chief Technologist Glenn Wightwick mentioned that an exabyte machine to cope with the data generated by SKA will be equivalent to 1 billion PCs.
Cloud and world data traffic
According to Cisco Global Cloud Index, global data center traffic is expected to quadruple from 2011 to 2016. Data will reach a staggering 6.6 ZB annually in 2016 from 1.8 ZB in 2011. This equates to nearly 1.5 years of continuous music streaming for the world’s 7.5 billion population in 2016. Interestingly, data center traffic in the cloud is forecasted to grow six-fold during the same period – from a 683 exabytes of annual cloud data traffic in 2011 to 4.3 zettabytes of data served by cloud in 2016. This tells us that around 65 percent of global data center traffic in 2016 will be contributed by cloud.
Figure 1 shows the distribution of data center traffic that we will be observing from 2011 to 2016, as per Cisco Global Cloud Index. User data (1.2 ZB) only contributes to 17 percent of the traffic.
Figure 1 Data center traffic pattern 2011 to 2016
Cloud will significantly dominate data center traffic as 62 percent (112 million cloud workloads) of all workloads will be cloud based by 2016. In year 2011, 30 percent (21 million cloud workloads) of workloads were processed in the cloud, whereas 70 percent were processed in a traditional data center. According to the forecast, in 2014, 52 percent of the workloads will be in cloud and the rest in traditional data centers. Distribution of workloads for 2011, 2014 and 2016 are shown in Figure 2.
This report also forecasted that average workload per physical cloud server will increase from 4.2 in 2011 to 8.5 by 2016, while average workload per physical server in traditional data centers will grow from 1.5 to two in the same period. Workloads are gradually finding a place in the cloud.
Figure 2 Workload distribution in cloud and traditional data centers
In the next chart, I will present cloud traffic growth by some regions.
Figure 3 Cloud traffic (in exabytes) by region
As we can see in Figure 3, Asia Pacific will dominate by producing 1500 exabytes or 1.5 ZB of data. According to Cisco’s forecast, the Middle East and Africa will have highest cloud traffic growth rate (79 percent CAGR). Second and third in the list are Latin America (66 percent CAGR) and Central and Eastern Europe (55 percent CAGR). It is evident from all this data that cloud adoption rate is dramatically improving.
Cloud is reshaping global data traffic and cannot be ignored. Businesses need to find out how to leverage cloud to grow their business and innovate.