Cloud Computing Week Jevin Review (4.9-4.14)

The market for large data practitioners: Pay continues to grow the talent gap is huge

In the big data age, businesses are fighting to attract and retain professionals in business intelligence and information management. Large data workers face a huge gap in the payroll survey of IT employees published every year in InformationWeek.

Today's large data presents a "4V + 1C" feature. Produced: generally includes the structure, the semi-structured and the unstructured and so on many kinds of data, moreover they processing and the analysis method to have the difference; Volume: through a variety of equipment produced a large number of data, PB level is the normal; Velocity: Requires fast processing, timeliness Vitality: analysis and processing models must change rapidly, as demand changes; complexity: processing and analysis are very difficult. and IM professional lets people know how to use the emerging large data platform (such as Hadoop, NoSQL) to process and analyze large data.

Reviewing InformationWeek's IT payroll survey reports for years, BI (Business FDI) and IM (information Management) professionals have been high income people in the IT field. The INFORMATION2012 survey showed that the average salary for BI practitioners was $90000, compared with a 2011 figure of $85000. The average salary for BI managers is $119000 trillion, up 9000 dollars over the same period last year. In the field of data consolidation/data warehousing, salaries were ever higher, with salaries for average practitioners and managers at $97000 trillion and 120000 dollars respectively, a small change from 2011 (The salaries of general practitioners and managers were 98000 and 118000 dollars respectively).

The average payroll for big data workers looks pretty good, but what more can be paid? The big data wave is pushing the CEO to find people who can tell him about future developments (nothing happened last week or even last month).

Such people have forward-looking and predictive insights. "They are usually statisticians and proficient in data modeling, and they know how to use the best algorithms in the available data, which is highly technical," said Stacy Blanchard, head of the cultural change group, a team of Accenture's talent and organizational performance service. They help organizations to exploit valuable data in a large amount of information and translate the data into models of in-depth cognitive and accurate predictions.

The big data age has also forced many industries to change, and today manufacturers are studying demand data and supply chain information in order to cut product development cycles and improve manufacturing and supply chain efficiency. At the same time to AOL, ComScore, eharmony level and the Internet companies are the customer's click Stream data analysis to provide valuable personalized and positioning services, while helping to find the most suitable customers for precision marketing.

From Starbucks to Wal-Mart, they are analyzing intensive data to produce the most appropriate product selection and pricing. for medical care. For example, the Harvard Medical College can improve the efficiency of diagnosis and treatment for patients by analyzing the clinical data. In short, companies and government agencies are increasingly focusing on data-driven decision-making, which leads to an increasing demand for analytics and information management expertise. In the future, the Big data field will have "salaries will continue to grow, large data talents in short supply, employment outsourcing, talent team divergence, large data professionals need continuous improvement, proficient in large data professionals will become the most important business role, large data fields need data scientists, universities respond to large data talent gap, Data-driven work is satisfying and challenging, and the big data professionals will embrace the future "10 of these trends."

Online video King YouTube's philosophy of technology

If you want to build a website that can host 4 billion of days of visits, YouTube has a lot to learn from. This article is a summary of YouTube's extended speech by Mike Solomon, a YouTube engineer, in Pycon (Pycon is the annual event for developers of the Python Open source community).

Many teams make their infrastructure increasingly complex, and the YouTube team tries to keep it simple. They use Python as their primary development language, use MySQL open source database, and always use Apache as a Web server. For a site as large as this one, many of the important features come from a bit of Python code.

This does not mean that YouTube does not pursue innovation, and that it pursues a philosophy of technology rather than hype. What is the biggest video site in the world? Please read.

New quantum encryption technology can thwart "most sophisticated" hackers

A research team led by Professor Rohai of the University of Toronto in Canada found that a new method of quantum cryptography could thwart the "most sophisticated" hackers. The findings are published in the latest issue of Physical Review Letters.

Quantum cryptography is, in principle, a foolproof way to stop hackers, it ensures that the disturbances caused by any attempt by the eavesdroppers to read encrypted communications are detected by legitimate users, and therefore, in the presence of a potential hacker, quantum cryptography allows the transmission of an absolutely secure key between two users. The key is transmitted using optical signals and is received by a photon detector. However, hackers can intercept and manipulate these signals.

Dr Charles Bennett, an IBM researcher and co-founder of Quantum cryptography, said that the photon detector proved to be the Achilles heel of key distribution, which inadvertently opened a door to the elusive "heretical" attack-the most famous quantum hacker behavior.

When quantum hacking occurs, light signals "subvert" the photon detectors, allowing them to see only the photons the hacker wants the other to see. In fact, Professor Luo's early studies and independent studies by Dr. Wadim Makarov of the Norwegian University of Science and Technology show that a clever quantum hacker can crack a business key distribution system.

What did you buy for 1 billion dollars? Instagram structure

Instagram is a social-networking image-sharing service on the iphone, which attracted 14 million users in the first year of its inception, and has now reached 30 million users. By the end of last August, Instagram had a total of 150 million images, all of which were based on Amazon stack.

Instagram's team once wrote "Instagram: Hundreds of examples of a large number of techniques" disclosed in its architecture.

Instagram mixes different technologies and strategies. The team is small, but the growing experience has allowed them to catch up with the social and mobile Internet bandwagon. They used mixed SQL and NoSQL technology, there is a large number of open source projects, and the choice of cloud services, Amazon's services provide a high leverage, which is much higher than their own construction, reliability is fully available, in a chronological order asynchronously link components, the system contains enough services, Includes APIs and external services that are not developed by engineers. The data is stored in memory and cloud, most code is dynamic language, the newly developed background transfer service links all the services together, and the code updates quickly and remains short. A very modern architecture.

IBM Design the strongest computer: throughput data equivalent to twice times the global Internet

According to foreign media reports, IBM is designing the most powerful computer in history, the number of daily data throughput equivalent to the entire Internet, can be used to analyze radio waves from space, explore the mysteries of the origins of the universe. The computer will be linked to an array of 1900 square miles (about 4920 square kilometres) of the array of square kilometres, which is responsible for analyzing the data obtained by the telescope. The square kilometer array was able to observe the cosmic scene 13 billion years ago, just before the Big Bang, which was the dawn of the universe.

IBM has designed the strongest computing power millions of times times the fastest PC, and is responsible for processing more than 100 times times the data produced by the Large Hadron Collider. "The square Kilometer array gets twice times more data per day than the current global internet traffic," says Tang-Emposen of IBM's research and development department. "It is reported that the square kilometer array will be completed in 2024, can observe the cosmic scene 13 billion years ago, will be used to study the galaxy evolution, dark matter and the origin of the universe."

IBM will build this computer with 3D chips to achieve the computing power required by the square kilometer array. The square kilometer array will be fitted with thousands of dish antennas covering an area of more than 1 square kilometres and the entire array is the same width as the continental United States. The telescope array is 50 times times more sensitive than any previous radio telescope, 10,000 times times faster than the current telescope. Each square kilometer array of data generated per square kilometer is estimated to reach a number of EB (ai bytes). After processing, the data stored annually is estimated to be between 300 and 1500PB (bytes). By contrast, the Large Hadron Collider of the European Nuclear Research organization produces about a year.

Eight reasons why cloud computing is more suitable for small companies

Cloud computing has reduced the cost of business for a wide range of companies, and the advantages of cloud computing are particularly evident in the small companies that have just started. But there is also the opposite view that cloud computing is only significant in large companies.

On the one hand, the economic benefits of cloud computing in the original budget is not much of the small companies clearly show no advantage, on the other hand, large companies cloud computing security, integration and other common problems in small companies are not serious. But the size of small companies, enterprise-class functions, costs, equipment and applications, compatibility, security issues, legal affairs, reliability and other eight aspects of the characteristics of the small companies are more suitable for the adoption of cloud computing.

Cloud computing seems inherently designed for small and medium-sized companies, and is almost the best way to reach an enterprise-level it level without having to develop it yourself or pay a large price for it. Although the technology is fledgling, the choice of cloud computing is not a smart move from the standpoint of building a new company.