Bangkok--4 Feb--Francom Asia
Huawei has recently launched OceanStor Dorado V6, an all-flash storage that sets new benchmarks in storage performance and reliability for enterprises' mission-critical service, while providing excellent user experience. With Huawei-developed intelligent chips, FlashLink(R) intelligent algorithm, and full-series end-to-end NVMe architecture, the OceanStor Dorado V6 delivers 20 million IOPS of best-in-class performance, twice that of the next-best player in the industry. The next-generation hardware platform and SmartMatrix architecture with a fully-interconnected design ensure always-on services. The AI chips are the first in the industry to make storage systems more and more intelligent during the application operations.
No.1 Performance with Chip-Powered Architectures
Equipped with intelligent chips, end-to-end NVMe architecture, FlashLink(R) algorithm, and powerful scalability of up to 32 controllers, the OceanStor Dorado V6 all-flash storage delivers 20,000,000 IOPS* of best-in-class performance and 0.1 ms of latency, twice that of the next-best player. OceanStor Dorado V6 is the perfect partner for enterprises that have data-hungry storage requirements of large databases, particularly in the carrier, finance, government and manufacturing industries.
Huawei all-flash storage uses five built-in intelligent chips for end-to-end application acceleration:
The intelligent multi-protocol interface chip hosts the protocol parsing previously performed by the general-purpose CPU, expediting the front-end access performance by 20%.
The intelligent processor chip sets new benchmark of performance, and its computing power is 25% better than the industry average.
The intelligent AI chip actively analyzes and understands the I/O rules of multiple application models based on machine learning frameworks to implement intelligent prefetching of memory resource. This improves the read cache hit ratio by 50%.
The intelligent SSD controller chip hosts the core Flash Translation Layer (FTL) algorithm, accelerating data access within SSDs and delivering a 10% less read latency than the industry average.
The intelligent BMC management chip has a built-in Huawei storage fault library, which speeds up fault location and diagnosis of components, and shortens the fault recovery time from 2 hours to 10 minutes.
Always-On Applications with SmartMatrix-Based 4-Layer Reliable Layouts
For customers that need robust IT systems that can consolidate multiple types of services and ensure stable service running, Huawei OceanStor Dorado V6 all-flash storage is the perfect choice. It ensures end-to-end reliability at levels ranging from component, product, solution to cloud layers, supporting the data consolidation scenarios with 99.9999% of service availability.
Empowered by an intelligent balancing algorithm and a self-developed convergent network chip, SmartMatrix architecture tolerates the collective failure of seven out of eight controllers across engines without causing service interruptions — a capability that no other vendor can provide.
Reliability: Toleration of the failure of seven out of eight controllers across engines without disrupting services. The unique SmartMatrix with the fully interconnected design at both front- and back-ends tolerates 7 controller failures without service disruption.
Availability: Only one second switchover with uninterrupted links in the event of controller failure LUNs have access to application servers through any controller rather than having ownership. Multiple controllers share loads running the load balancing algorithm. If a controller fails, other controllers can smoothly take over services without any interruption or performance loss.
Serviceability: Only one second online upgrades with no impact on host services. It applies the 3D fixed structure, the combination of vibration isolation and absorption, and the damping screws with non-rigid material, to ensure always-on services in the event of severe vibration. It is the industry's first to pass the magnitude-9.0 earthquake resistance test by the China Telecommunication Technology Labs (CTTL).
Efficient O&M with Edge-Cloud AI Synergy
Based on AI chips and AI algorithms, the OceanStor Dorado V6 all-flash storage implements intelligent O&M throughout the entire lifecycle. The innovative business model avoids service interruption caused by data migration, providing smooth user experience.
Edge-cloud synergy: General-purpose cloud AI, customized edge AI, and built-in Ascend 310 chips are used for incremental training and deep learning of service characteristics, enhancing personalized customer experience
AI throughout service lifecycle: Intelligent management enables 60-day prediction of performance and capacity trends, 14-day prediction of dish faults, and offers immediate solutions for 93% of problems located.
Flash Ever program: The intelligent elastic architecture implements component-based upgrades without the need for data migration within 10 years. Users can enjoy latest-generation software and hardware capabilities, while also making the most of their existing investments.