Recently, Inspur information released a new generation of SSD high-speed storage media. Based on NAND algorithm innovation, this new product will increase the life of flash memory by 40%. While realizing 1.5 million IOPs per disk through PCIe 4.0 ultra wide channel and ZnS storage technology, it also helps Inspur storage to win the first in spc-1 performance in the world, and build core competitiveness in reliability, performance and system joint commissioning optimization.
The report shows that the digital economy and GDP will grow by 3.5 ‰ and 1.8 ‰ respectively for every point of increase in the computing power index. At the same time, the data scale is expanding rapidly, with a compound annual growth rate of about 26%. By 2025, the total amount of data creation will increase significantly to 175zb. Now, more data are created per hour than in the past 20 years.
In order to create a core carrier for data elements, Inspur has continuously penetrated into the underlying technology and created a new generation of SSD high-speed storage media. Inspur new generation SSDs perform well in terms of reliability and performance. More importantly, they can provide full chain optimization from protocol to storage media.
First of all, security and reliability are the basic requirements of enterprises for storage media. It can even be said that for key businesses, data is the “lifeline” of enterprises.
Inspur SSD adopts an innovative NAND characteristic management algorithm and AI intelligent management algorithm to detect NAND parameters and submit them all to the platform. Through platform modeling and optimization, the best read voltage is obtained, thus improving the reliability of the disk. The mean time between failure (MTBF) reaches 2.6 million hours.
Secondly, driven by the global green trend, the data center bears a dual mission: on the one hand, with the help of diversified technical means, reduce its own energy consumption and reduce pue to a level below 1.3 or even lower; Second, improve data processing efficiency. Inspur SSD can reduce power consumption while pursuing extreme performance.
Inspur SSD adopts an 8-level power consumption dynamic adjustment mechanism to achieve W-level power consumption adjustment. It is constantly innovating in NAND media applications, and its energy efficiency is more than 70% higher than that of the previous generation; At the same time, Inspur has greatly improved the performance by widening the channel between NAND and controller. The random reading performance of the new generation PCIe 4.0 SSD is up to 1.5 million IOPs. What is particularly important is that Inspur SSD can cooperate with the storage system to optimize. By optimizing the read-write ratio model of various scenarios, it can fully open up the IO path of the storage system and SSD disk, which can improve the overall performance of the system by 30%. Inspur SSD has won the first place in the world in terms of storage performance in spc-1 international benchmark, forming a highly competitive product system.
Inspur combines the cutting-edge technologies such as ZnS storage and kV key value of SSD with the public cloud scenario, so that the whole machine can better understand SSD, make SSD more suitable for solutions, and greatly improve the public cloud user access experience. At present, Inspur enterprise SSD has been deployed on a large scale in finance, Internet, communication and other industries. According to the latest Gartner data, Inspur’s storage sales leapt to the top four in the world, becoming the world’s strongest growth manufacturer.
With the emergence of new applications such as AI, big data and meta universe, big data center, as a new infrastructure in all walks of life, is experiencing rapid development. Inspur has realized the research and development of full stack storage from device components to the whole machine, from hardware architecture to core software stack and management software stack. In the future, Inspur storage will work with partners to accelerate the construction of scene community and release data value.