HITACHI Advanced Server DS7000 Series Scalable Servers

HITACHI Advanced Server DS7000 Series Scalable Servers

Overview

In the wave of digital transformation, enterprises need to leverage the latest technologies such as artificial intelligence, data analysis, and machine learning to maintain competitiveness. This requires the underlying infrastructure to have extremely high reliability, ultimate performance, and agile scalability. The Hitachi Advanced Server DS7000 series scalable servers were born for this purpose, featuring a unique modular architecture designed to meet the most demanding IT challenges and provide powerful power for critical business applications, including in memory database environments, artificial intelligence, and machine learning.

The core design concept of the DS7000 series servers is based on a universal computing module built on two Intel Xeon Scalable processors. This design enables smooth upgrades between various models from DS7020 to DS7080 and even DS7160, maximizing the protection of enterprise investments in hardware and software, and simplifying the complexity of system reconstruction and expansion.

Core Architecture: The Revolution of Modular Design

The cornerstone of the DS7000 series is its innovative modular architecture. Each computing module is an independent computing unit, containing two Intel Xeon Scalable processors, corresponding memory slots, and I/O capabilities. This' building block 'style approach brings unprecedented flexibility:

On demand expansion and smooth upgrade: Enterprises can choose the entry-level DS7020 (2U, 2 computing modules) according to their current business needs. When the business grows and requires stronger computing power, it can be upgraded to DS7040 (4U, 4 computing modules) or DS7080 (8U, 8 computing modules), or even the top-level DS7160 (21U, 16 computing modules). This expansion is not just a simple stacking, but a seamless integration of internal interconnection and resource pooling within the system.

Resource pooling and flexible configuration: Each computing module can be independently configured to support multiple internal computing and storage options. This means that different specifications of processors, memory, and storage devices can be mixed and configured within a DS7000 chassis to meet the personalized needs of different application workloads, such as configuring high-capacity memory for memory databases and high-performance I/O for data intensive applications.

Hardware partitioning support: Starting from DS7040, the DS7000 series supports hardware partitioning. This allows a physically large server to be divided into multiple independent, electrically isolated logical partitions, each running a separate operating system and application, achieving true resource isolation and integration, making it an ideal choice for integrating critical business workloads.

Technical specifications and performance

The DS7000 series offers multiple models ranging from 2U to 21U to meet business needs of different scales. All models are based on the Intel Xeon C627 chipset and share the same RAS (Reliability, Availability, Serviceability) design philosophy.

3.1 Overview of Product Series

DS7020 (2U): An entry-level scalable server that supports 2 processors and up to 3TB of memory. It is applicable to small and medium-sized databases, virtualization portals, and edge computing scenarios. Provides up to 7 PCIe Gen3 slots and 8 2.5-inch front hot swappable hard drives.

DS7040 (4U): Mid end flagship model, supporting 4 processors and up to 6TB of memory. It is an ideal platform for medium-sized virtualization clusters, in memory databases, and business intelligence applications. Provides up to 14 PCIe Gen3 slots and 16 2.5-inch front facing hard drives.

DS7080 (8U): A high-end enterprise grade server that supports 8 processors and up to 12TB of memory. Specially designed for large-scale in memory databases (such as SAP HANA), high-performance computing, and large-scale virtualization consolidation. Provides up to 28 PCIe Gen3 slots and 32 2.5-inch front facing hard drives.

DS7160 (21U): Flagship model, supporting 16 processors and up to 24TB of memory. It is a 'performance monster' designed for the most demanding critical business workloads, providing unparalleled computing density and scalability. Its I/O capability is equally impressive, supporting up to 40 PCIe Gen3 slots and 64 2.5-inch front facing hard drives.

3.2 Powerful I/O and Storage Scalability

The DS7000 series not only has powerful computing power, but also excellent I/O and storage scalability.

PCIe Expansion: All models are equipped with a wide range of PCIe Gen3 slots, which can be used to install high-speed network adapters, storage host bus adapters (HBAs), and GPU acceleration cards. For example, the DS7080 supports up to 8 GPUs, providing powerful acceleration capabilities for AI and deep learning workloads.

Storage flexibility: Except for the front hard drive, all models can be equipped with an optional "upper unit" to expand more 2.5-inch or 3.5-inch hard drives. When fully configured, DS7160 can support up to 96 2.5-inch hard drives or 32 3.5-inch hard drives, providing massive local storage space.

Network connection: The standard configuration provides 10Gb Ethernet ports to ensure the bandwidth requirements for data transmission. Meanwhile, users can flexibly choose 10GbE network adapters with 1GbE or more ports through PCIe modules.

Enterprise level RAS features: the cornerstone of reliability

Reliability, availability, and serviceability (RAS) are crucial for servers running critical business applications. The DS7000 series has integrated RAS features into every component and level since its design, ensuring that the system can run continuously and stably 24/7.

4.1 Advanced Error Detection and Correction (AEDC)

The DS7000 series adopts a multi-level and comprehensive error detection and correction mechanism, covering all critical paths from processor interconnect, memory to PCIe bus.

Memory protection: Supports DDR4 VM driver check/retry, command/address parity and retry, as well as fault DIMM isolation technology. When a memory error is detected, the system can automatically isolate the faulty memory area to prevent the entire system from crashing.

Processor interconnect protection: Intel UltraPath Interconnect (UPI) link layer supports retry and dynamic link width reduction. Protect the protocol through a 16 bit rolling CRC to ensure the reliability of communication between multiple processors.

PCIe bus protection: supports PCIe ECRC, data poisoning, link CRC error checking and retry, link retraining and recovery, and real-time error recovery. When there is a problem with PCIe devices or links, the system can attempt to recover to avoid business interruption.

4.2 Fault isolation and system recovery

Viral error isolation: The system can effectively prevent individual errors from spreading within the system, isolate them within faulty components, and ensure that other parts are not affected.

PCIe 'Stop and Scream': A fast fault notification mechanism that can immediately notify the system when a fatal error occurs in a PCIe device, allowing for quick protective measures to be taken.

Fault elastic startup: supports disabling the faulty core during the startup process to ensure that the system can still start and run normally.

Post encapsulation repair: allows for the repair or isolation of certain encapsulation level faults after the system is running.

4.3 Excellent Serviceability Design

In order to minimize unplanned downtime to the greatest extent possible, the DS7000 series has been designed with maintenance convenience in mind.

Fully modular hot swappable design: The power supply, fan, hard drive, and some PCIe modules supported by the operating system all support hot swappable. This means that faulty components can be replaced without shutting down the system.

Pre maintenance design: The hard drive and some key components adopt a pre access design, and maintenance personnel can perform maintenance operations without pulling the server out of the cabinet.

Redundancy Design: All critical subsystems, including power supply (1+1 redundancy/computing module), fans (N+1 redundancy), and hard drives (RAID configuration), are designed with redundancy to eliminate single points of failure.

Born for critical business solutions

The outstanding performance and reliability of the DS7000 series make it an ideal platform for running critical business solutions.

5.1 Hitachi SAP HANA Platform Solution

SAP HANA is an in memory database platform that requires extremely high memory capacity and performance from servers. The DS7000 series, especially the DS7080 and DS7160, with their memory capacity of several terabytes and powerful processor performance, can fully utilize the in memory computing power of SAP HANA to help enterprises obtain real-time business insights from massive data and accelerate decision-making processes.

5.2 Hitachi Database Solution

For critical database applications such as Oracle Database, I/O performance and system stability are crucial. The DS7000 series, with its powerful CPU processing power, massive memory cache, and local storage connected via high-speed PCIe bus, can provide enormous I/O throughput capability. This ensures that enterprises can meet stringent service level agreements and cope with rapidly changing workloads with ease, ensuring the flourishing development of the business.

Summary

The Hitachi Advanced Server DS7000 series scalable server is an infrastructure platform designed specifically for the digital age. It achieves smooth expansion from 2U to 21U through innovative modular architecture, meeting the diverse needs of enterprises from small applications to large-scale critical business workloads. Its enterprise level RAS features throughout ensure stable operation of the system in harsh environments. Whether as an in memory computing engine for the SAP HANA platform or a reliable cornerstone for large databases, the DS7000 series provides unparalleled performance, availability, and investment protection, serving as a powerful driving force for enterprise digital transformation.