[Beijing, China, April 3, 2019] At today’s +Intelligence, Taking Computing to New Levels press conference, Huawei Intelligent Computing Business Dept announced the FusionServer Pro series of intelligent servers. These next-generation x86 servers accelerate the intelligent transformation of data centers thanks to an intelligent acceleration engine, intelligent management engine, and intelligent data center solutions for diverse scenarios.
Huawei FusionServer Pro launch ceremony
Transformation today is shifting from digital to intelligent. As the core driving force, the computing industry faces challenges in compute power and management. Featuring outstanding compute power and intelligence, Huawei’s x86 FusionServer Pro will harness the computing industry and enable transformation into an intelligent world.
This intelligent world needs infrastructure consisting of three parts: connectivity, compute, and cloud. Connectivity and compute infrastructure make the cloud more efficient. And data centers are responsible for most computing tasks. According to Huawei’s Global Industry Vision (GIV) 2025, the penetration rate of Artificial Intelligence (AI) in enterprises will reach 86 percent by 2025. Along with increasing intelligent transformation in industry, more and more large-scale data centers will be built. Traditional data centers and servers behind them must upgrade to address the following existing challenges:
High costs of traditional servers: In recent years, new services are emerging all the time. The traditional mode of using only CPUs is not the most economical choice in certain service scenarios.
High OPEX of data centers: It’s costly to deploy, manage, and maintain a large number of servers in traditional data centers.
Lack of L1-L2 linkage: In traditional servers, air conditioners at L1 and servers at L2 are not linked. As a result, the overall energy efficiency is low, and the electricity used by air conditioners costs up to 30% of total data center electricity usage.
To address the intelligent transformation requirements of data centers, Huawei upgrades its FusionServer series to FusionServer Pro intelligent servers and provides tiered solutions covering single-node, large-scale, and ultra-large-scale deployment scenarios.
Single-node deployment: The intelligent acceleration engine implements intelligent offloading for computing, storage, and network protocols to improve application speeds and service experience, reduce operation costs, and increase return on investment (ROI).
Large-scale deployment: The intelligent acceleration and management engines implement intelligent asset, version, power consumption, deployment, and fault management, improving management and O&M efficiency and reducing OPEX by 15%.
Ultra-large-scale deployment: Besides the intelligent acceleration and management engines at the device level, three innovative intelligent data center solutions are provided:
Intelligent L1-L2 linkage: This solution links the air conditioners at L1 and the servers at L2 to improve the overall energy efficiency of data centers.
Full liquid-cooling cabinet solution: This energy-saving solution can reduce Power Usage Effectiveness (PUE) by 30% for data centers. Its high-density deployment can save 76% of cabinet and 68% of equipment room space. Compared with the air cooling solution, the solution can reduce total cost of ownership by 20%.
Integrated cabinet solution: This solution introduces large-scale and rapid deployment, node blind-mate insertion, and intelligent robot O&M to make unattended intelligent data centers a reality.
Zhang Xiaohua, Huawei Intelligent Computing Business Dept, giving a speech
FusionServer president Zhang Xiaohua noted: “By introducing an intelligent acceleration engine and an intelligent management engine, Huawei upgrades its traditional servers to intelligent servers to help enterprises build intelligent data center infrastructure with unprecedented compute power.”
For more information about Huawei Intelligent Computing products and solutions, visit: e.huawei.com/en/solutions/hic.
PR Archives: Latest, By Company, By Date