All Australian Company

Sales: 8am to 8pm 7 days

PH : 1300 00 1000

High Performance Compute Servers/Lenovo: Lenovo, SD530, dual, Node, System, with, D2, chassis., Each, Node, has, dual, 6254, (36, cores/3.1ghz), 256GB, RAM, 2*1.92TB, Solid, Stat,

Lenovo SD530 dual Node System with D2 chassis. Each Node has dual 6254 (36 cores/3.1ghz), 256GB RAM, 2*1.92TB Solid Stat

Our Price (inc. GST):

$166,000

$3,515.89 /month*

Freight Cost:

Syd: 

$88.62

Melb: 

$126.39

Calculate freight to your postcode:

Full Australian Warranty

Australian Distributors

Product is in a sealed box

PRODUCT ID: SD530HPC1

Description

SD530 dual Node System with D2 enclosure. Each node has dual 6254 (36 cores at 3.1ghz), 256GB RAM, dual 1.92TB Solid State Drive (SSD), dual V100 32GB

This is a 2U D2 enclosure with dual SD530 nodes. Each node supports dual double-width GPUs.
Each node has dual 6254 processors providing 36 cores at 3.1ghz. Each node has 256GB RAM, dual 1.92TB SSDs, dual 10GBE and dual double-wide V100 32GB GPUs.
This provides a total of 72 cores at 3.1GHZ, 512GB RAM, 4 * 1.92TB SAS SSDs, 4 * 10GBE ports and four (4) Nvidia V100 32Gb cards (total 20480 CUDA cores and 2560 tensor cores) in a 2U form factor.
The D2 enclosure features 16 memory slots, 10GBE ports, up to 12 drive bays (max 6 per node) and dual 2000W power supplies. This enclosure has two GPU trays fitted to accommodate the four V100 32GB GPUs and is limited to two nodes. The D2 Enclosure requires 15A power due to the large power supplies supporting the GPUs.



The Lenovo ThinkSystem SD530 is designed to excel not only at critical enterprise workload environments (such as virtualization, hyperconverged infrastructure, and cloud), but also at High Performance Computing (HPC) and Artificial Intelligence (AI)

The V100 32GB GPU is Nvidia's highest powered GPU with 5120 CUDA cores and 640 Tensor cores for Analytics, Machine Learning and Deep learning. NVidia claim it is the Most Advanced Data Center GPU ever built.The dual cards may be shared over multiple virtual machines as virtual GPUs (vGPUs) or linked to leverage more than 10000 available cores for a single workload.

See: https://www.nvidia.com/en-au/data-center/v100/

This system can leverage Nvidia's NGC library with downloadable containers, models and scripts to leverage these GPUs and get your AI initiative yielding results much quicker.


The system can be supplied with VMWare ESXI, Nvidia's vComputeServer and even Docker and NGC containers preloaded. Nvidia's vComputeServer allows for virtual GPUs to be assigned to different virtual machines. Depending upon the workload it is possible to aggregrate physical GPUs or split physcial GPUs.


All SD530 High Performance Computing nodes are supplied as special order with a normal leadtime of 4 weeks. This ensures that the latest hardware revisions and firmware is supplied ex-factory.

Alternate processor, storage, RAM and network configurations are available. Please contact us to discuss your specific requirements.

Technical Specification

Processors: Each node has Dual Intel 6254 18 core processors (36 cores at 3.1GHZ). Alternate Gen 2 processors available.
RAM: 256GB RAM ( 2666mhz 32GB RDIMM * 8) pre-installed. Support for up to 12 or 16 RDIMMS depending upon model.
Array Controller: Embedded Intel RSTe software RAID, supporting RAID 0, 1, 10, 5, 50
Storage: 2 * 1.92TB 6gbpsSSD
Network: 8 port 10GBE installed in D2 enclosure. 2 ports available for each node.
Power: Dual 2000W power supplies in D2 Enclosure.
Cooling: 5 hot-swap system fans
System Management: Integrated service processor (XClarity Controller (XCC)) which provides advanced control, monitoring, and alerting functions. System Management Module (SMM) is the management device for the enclosure. Provides integrated systems access via the dedicated Gigabit Ethernet port. Remote access is to both the management functions of the enclosure as well as the XClarity Controller (XCC) in each server.

GPUs: Each node has Dual Nvidia V100 32GB GPU pre-installed. This system is fully populated with 4 V100 32GB GPUs.
Operating system: RedHat, SUSE and Ununtu certified. Windows Server 2019 and Windows Server 2016
Virtualisation: Support for ESXi and Hyper-V. ESXi required for Nvidia vComputeServer vGPU technology

Warranty

A 3 year warranty with 4 hour response and 24 * 7 service is included with this server.

Last updated: Friday, 29 March 2024

Warranty

This product is sourced in Australia and carries a full Australian Manufacturers Warranty

Brochure

Brochure