Data center migration, now, data center has become the cornerstone of today's global economic development. They can move, store and analyze information that is located in the center of personal and commercial life. Without data centers, people can't communicate freely, share social platforms, watch streaming videos, do business activities, acquire and transfer funds freely. There is no better application of artificial intelligence, machine learning, Internet of things, security, 5G network, autonomous traffic and other emerging technologies.
According to the Research Institute, the auto - driving camera, radar, sonar, GPS and laser radar system will produce 4TB data every day. If 10% of today's cars in the United States are autopilot, 40 ZB of data will be generated each year. It is expected that there will be 30 billion or more Internet of things equipment in the world by 2020. The Internet of things devices will generate and use 44ZB byte data, and the processing of these data requires 400 million servers in 4000 large data centers. The development of artificial intelligence, machine learning, speech recognition and image recognition will open the door to a large number of unprecedented information. These information demand greater demand for the cutting-edge ultra large scale data centers.
In 2017, about 8 million data centers around the world (from small server cabinets to large data centers) were dealing with data loads. These data centers consumed 416.2 TWH (1 terawatts of approximately 1 billion kwh of electricity). This is equivalent to 2% of the world's total electricity consumption and is expected to reach 5% of the world's electricity consumption by 2020.
It is clear that most of the load and data of modern life are fast or will be completely dependent on the data center. But with the growth of data traffic, a large number of data center construction, people in through the data to create a better world at the same time, the data center will face growing resources and power demand, need more energy, steel, concrete, copper production data center power every day.
If people no longer rethink the efficiency and indicators of the data center, they may need to pay a heavy bill in the near future.
Data center data
The development of data center market not only changes with the increase of data volume, but also increases with time, security and IT cost. The result is a series of complex services, infrastructure, operating expenses and business models for the organization. Many organizations move their businesses running in small server rooms to managed cloud solutions, and businesses, multi tenants, super large scale manufacturers, and high-performance computing data centers now occupy the largest share of growth and energy consumption.
The average power consumed by the data center is more than 100 times more than that of large commercial office buildings. A large data center uses the total amount of electricity equivalent to a small American city. The energy required for the power supply and cooling of the server accounts for 40% of the total operating cost of the data center, so the energy inefficient problem of the data center should not be ignored.
These problems are well known, and effective solutions must be widely adopted.
Data center power inefficient and power density problems will become an important business issue for enterprises' Sustainable Development Directors and their "green / sustainable development" teams.
People need to put forward new problems and indicators:
For the head of the data center: how big is the data center's energy bill?
For CFO, how much money will be needed to build the data center, or can it be delayed? How much revenue can an enterprise get through a data center device? Should enterprises own and operate their own data centers or should they use a managed data center service?
For the chief executive, how can the enterprise achieve greater success in the near future and in the near future?
Besides the industry leaders of super large or high-performance data centers (such as apple, Facebook, Google, Amazon, Microsoft, Baidu), enterprises can understand these problems, but these solutions have not been implemented to a large extent. Among the various concerns of the data center operators, the priority of energy costs is usually not high.
There are three main strategies below.
The 1. data center facilities efficiency and software technology are well documented.
2. renewable energy plan.
3. most undeveloped hardware solutions.
The data center can use these three main strategies to solve power inefficient and density problems, and also need to change the corporate culture and management practice.
沒有留言:
張貼留言