Eight data center technologies verified in 2015

2015 was a year in which new technologies matured. The gradual commercialization of IoE, hybrid cloud and big data technologies made data centers more efficient.

In 2014, the market had new developments in business budgets. New technologies such as cloud platforms, big data, and advanced analytical methods enabled the business market to find profit points again.

[[126582]]

So, what will 2015 bring? Here are eight data center technology innovations that are sure to make the most of your data center.

1. Converged system. Customized servers based on needs. With the increasingly detailed division of labor in IT, the traditional unified server system can no longer meet the needs of all data centers. Some data centers may require high-density parallel computing, so they need to expand a large number of GPU integrated computing, and traditional servers obviously do not consider this; some data centers are mainly for data storage, which means that data centers with special needs such as storage expansion can freely match servers with more storage, including expandable NAS, etc.

2. Mesh network. The mesh network is what we mentioned in the previous article. The existing east-west and north-south networks have performance bottlenecks, which can be improved by adding an intermediate switching layer. Most companies still use hierarchical networks because they support users to access specific applications in a north-south network traffic pattern. The flattened network topology of the mesh network will improve east-west and north-south communications.

3. Flash media storage. The computing bottleneck in data centers is often not the CPU, memory or network speed, but storage. The storage performance has improved significantly, the most obvious of which is the use of Flash media storage, which includes SSD, SSHD, hybrid accelerated SAN and other technologies. FLASH media storage is the future development trend. Although FLASH has its own shortcomings compared with traditional disk storage in terms of capacity and fault recovery, the progress of technology is huge, and FLASH storage will soon enter the data center on a large scale.

4. Hybrid computing. Different tasks require different resources. The computing power of traditional x86 is limited in the face of the ever-expanding data computing. The CPU computing can no longer meet the needs of large-scale single parallel computing. Graphics processing units (GPUs) and APUs from companies like Nvidia and AMD, or Java offload engines, will partially replace the computing work of the CPU architecture, such as parallel encoding, transcoding, etc. These are much more efficient than x86 CPU computing.

5. Hybrid cloud technology. 2015 is expected to be the year of large-scale expansion of hybrid cloud technology. The flexibility of hybrid cloud technology can greatly improve workload. With the popularity of virtualization and hybrid computing, the choice between software and hardware products of different manufacturers is getting bigger and bigger. Hybrid cloud technology is to choose what is needed and accommodate expansion. In the data center, how to expand hybrid cloud technology will be the focus of future development.

6. Internet of Things (IOE). With the rapid growth of data volume, household devices or any other devices are becoming more and more intelligent, even including air conditioners, washing machines and other equipment. The Internet of Things will be an upgraded application of all the basic intelligent devices. For smart buildings, automated workshops, advanced tracking and customer analysis, the Internet of Things is crucial.

The Internet of Things will collect massive amounts of data. Although these big data may seem messy, they are the basis for commercial big data analysis, which is also a factor in the large-scale expansion of the Internet of Things.

7. Improve energy efficiency. From the perspective of green energy conservation, sustainable energy and improving energy conversion efficiency are goals that every company should have.

The cooling method for modern data centers uses as little energy as possible. In non-intensive computing environments, the server processors can consider using low-power processors from the perspective of performance ratio. These processors often reduce power consumption by 50% compared to ordinary processors of the same grade, which will be of great help in server heat dissipation and improving performance ratio. Similarly, the energy consumption ratio also needs to be considered when selecting components such as the power supply of the equipment.

8. Alignment with the business. Better understanding of what the business is will ensure that the business is viable and effective.

From a business perspective, finding out the best constraints and recommendations for projects and goals will be an important part of decision making. As the world changes, business has been considered as the core part of an enterprise.