Full Stack Data Center

Renew Computing & Storage

Moore's Law will gradually become obsolete. Single computing power will soon be unable to meet diversified service requirements. It is therefore clear: computing power and storage must be reconstructed, in order to overcome performance bottlenecks.

All-scenario flash storage — powered by a Non-Volatile Memory express (NVMe) Over Fabric Plus (NoF+) network — achieves data processing speeds twice those of the industry average.

xPU synergy in hyper-converged infrastructure improves Virtual Machine (VM) density by 30%.

Renew Network

With the scale of network services growing, efficiency has to keep up. From 2019 to 2021, east-west network traffic, for example, increased by over 50%, year-on-year. This has impacted network configuration and management, which has become increasingly complex as data center resources grow. This is true for both overlay and underlay networks, soft and hard Software-Defined Networking (SDN), Fibre Channel (FC), InfiniBand (IB), and Internet Protocol (IP) networks.

Network capabilities now need to be renewed to improve network agility and transmission efficiency.

Data-network synergy shortens service provisioning time across Data Center Networks (DCNs) from months to minutes, guaranteeing Service Level Agreements (SLAs).

Lossless IP and optical transmission networks, offering the industry's largest bandwidth, break the bottleneck of network transmission between data center cores. For instance, 400G Ethernet provides 10 times more bandwidth than FC.

Renew Reliability

With extra emphasis now being placed on business continuity, data center architecture is moving from standalone to geo-redundant and even multi-location and multi-active. With diversified deployment environments, Disaster Recovery (DR) and backup has become far more complex.
In simple terms, data centers must be overhauled to support high availability for services: anytime, anywhere, and across all-layers.

In simple terms, data centers must be overhauled to support high availability for services: anytime, anywhere, and across all-layers.

All-layer DR: The Recovery Point Objective (RPO) must now equal zero. The Recovery Time Objective (RTO) must be less than or equal to zero. Serial Optical Channel Converter (SOCC) capability ensures storage Input/Output (I/O) switchover falls from minutes to just seconds.

Full-cycle backup: Improves backup efficiency and maximizes the reuse of on-premises data on the cloud.

Unified management: One platform reduces the skill level requirements of Information Technology (IT) staff. Switchover is precise and efficient, ensuring business continuity.

Renew Green

Traditionally, data centers need power — lots of it. Indeed, collectively they account for 2% of the world's total electricity consumption. No surprise, then, that 10-year electricity fees represent 60% of the Total Cost of Ownership (TCO) of a data center.. However, it doesn't have to be like this. Huawei's solution renews data center facilities, optimizing energy consumption and achieving significantly more efficient Power Usage Effectiveness (PUE) through green contrscture, green power supply, green cooling, and green smart management, ultimately helping to build far greener, low-carbon data centers.