Five Highlights of 4U ThinkSystem SR860 V2 Server From Lenovo

By Randall Lundin, mission critical product manager, Lenovo Data Center Group

The Lenovo ThinkSystem SR860 V2 is a 4U server capable of supporting up to 4 processors and 48 HDDs or SSDs.

It has highlights in areas of high memory, scalability, storage, GPUs and reliability.

This article introduces 5 key aspects of the new server. It is designed for your most demanding, mission-critical workloads, such as in-memory databases, large transactional databases, AI, real-time analytics, ERP, CRM, and virtualized server workloads. Let’s take a look at 5 highlights of the SR860 V2 server.

Scalability to grow
Scalability allows a system to expand to meet increasing performance and sizing workload demands. Scaling memory is how much memory you can add to your server. Processor scaling is how many and the types of processors you can add to your server. Enterprises need assurance that the architecture is designed in such a way that system capacity can be increased as the needs of the business change.

The SR860 V2 provides scalability to meet increasing workload demands and allow to scale the system to meet your needs. It supports from 2 to 4 processors, expanding being is simple as docking the top processor and memory tray into the server.

In terms of storage, it supports up to 48 2.5” HDD/SSD and up to 24 NVMe drives. For rear I/O, you can add PCIe risers to increase I/O capabilities up to 14 rear facing PCIe 3.0 adapters plus a dedicated OCP 3.0 LOM. It supports one DIMM per CPU or up to up to 12 DIMMs per CPU to match memory needs. GPU capabilities range from single GPU all the way to 8 GPUs.

Internal view of the SR860 V2 (GPU trays removed)

Storage capabilities
The SR860 V2 provides up to 48 2.5” storage bays. Within these bays, you can install up to 48 SAS/SATA HDDs or SSDs. The system also supports two 7mm rear accessible drives used for boot drives. In term of NVMe, it supports up to 24 NVMe drives which is more than double the number of NVMe drives that the predecessor system supported. With the NVMe drives the SR860 V2 supports Virtual RAID on CPU (VROC) which is an enterprise, hybrid RAID solution designed for NVMe SSDs.

Advantages of onboard local storage:

Speed of accessing your data. Local storage results in low latency and fast response times. No worry of upload or download speeds retrieving your data.
Better security and control. Keeps critical data in-house. No third party has access to your information. Healthcare is especially sensitive to who and where the data is stored.
Control over the data because the data is stored locally and you have control over the server.
The data is easily accessible.
Less regulatory hurdles. Some countries require to process and store data within the country.
No competing for storage resources with shared storage.
Saves on cost and rack space without the need to purchase separate storage chassis.

ThinkSystem SR650 V2 with 24 AnyBay bays and 24 SAS/SATA bays

Accelerate workloads with GPUs
The SR860 V2 has support for up to four double-wide Nvidia V100S GPUs or up to eight single-wide Nvidia T4 GPUs to tackle compute intensive applications accelerating deep or machine learning, AI, training, analytics, 3D modeling, video transcoding and virtual deck tops.

SR860 V2 with GPUs can transform business and help solve challenges.

Allows researchers to analyze large volumes of data for rapid insights where simulation alone cannot predict the real world in areas such as bioinformatics, drug discovery and high energy physics.
Boost the throughput of data centers with fewer nodes, completing more jobs and improving data center efficiency.
Ensures VDI users enjoy the same quality virtual desktop experience from anywhere on any device. From knowledge workers and mobile professionals to designers and engineers it accelerates the VDI experience delivering better performance than CPU-only VDI environments.

Internal view of the SR860 V2 with GPU trays installed

High memory and virtualization capability
The Lenovo SR860 V2 provides high memory and virtualization capabilities.

A server with In-memory computing allows the processing of tremendous quantities of data in main memory to provide immediate results from analysis and transaction. The data is available for processing or analysis immediately after it is created.

Virtualization capabilities create a more dynamic and fluid environment in which compute capacity can be treated as a resource pool, protecting application service levels and improving agility and responsiveness to business needs. Importantly, virtualization isolates workloads so that they do not interfere with one another.

The SR860 V2 with the 3rd Generation Xeon Scalable processors supports up to 12TB of memory. This amount of memory enables it to support memory intensive workloads and implement large vVMs or run large in-memory databases without compromise such as SAP HANA.

Intel Optane Persistent Memory 200 Series module (top) and standard memory DIMM (bottom)

The SR860 V2 supports Optane Persistent Memory 200 Series which is the next gen persistent memory modules that enable memory capacity and fast access to persistently stored data. Unlike conventional DDR memory, persistent memory retains its data even if power is unexpectedly lost and provides faster access to persistent data than reading from a mainstream SSD.

Reliability and availability for mission critical workloads
Mission-critical workloads support the business in every dimension – its organization, its employees and its customers. These workloads are the demanding types of applications and databases running on servers, and include ERP, CRM, data warehouses, business intelligence and analytics. They must be available on a 24 x7x365 basis.

Any outage of the mission-critical server infrastructure results in reduced revenue, profitability, employee productivity, and customer loyalty. Any interruption in the operation or availability of these workloads will have a ripple effect throughout the organization because downtime interferes with business processes and interrupts business continuity. That means that mission-critical workloads must run on reliable and available platforms to ensure that business processes are running smoothly.

Lenovo servers have high customer reliability rating. The company’s x86 servers had the best uptime among all x86 platforms for the 7th straight year.

The SR860 V2 features reliability, availability and serviceability (RAS) features. Including Intel ‘Run Sure’ RAS features and Lenovo RAS features. It provides detection and self-healing of errors as well as the ability of the system to continue processing production workloads, even in the presence of faults. The self-healing technology proactively identifies potential failures and transparently takes necessary corrective actions. Light path diagnostics allows to diagnose the source of problems in seconds instead of hours or days. It works hand-in-hand with Predictive Failure Analysis (PFA), which monitors multiple types of system components within the server.

For system status, firmware, network, and health information, the SR860 V2 supports the typical LEDs for power, network and errors, and also has a LCD panel that allows to scroll up and down to access different menu options.

ThinkSystem SR860 V2 with integrated diagnostics display open
Click to enlarge

LCD displays following type of information:

System errors such as fan failure, power supply failure, PCI error, etc.
System VPD information (such as machine type and serial number)
System firmware levels such as UEFI and XCC code levels
XCC network information such as MAC address and IP address
System environmental information such as temperature, voltage, power consumption