AMAX ClusterMax Apex CPU-intensive, high performance HPC cluster

ClusterMax® Apex

HPC Cluster Solution for CPU-Intensive, Large Scale Deployments

 

  • High density balanced compute solution for large scale power conscious users
  • Based on new Intel Sandy Bridge architecture for best in class performance.
  • Integrated web based cluster management for turnkey provisioning and hassle free management
  • Optimized MPI implementations and pre tuned networking components to get parallel application up and running
  • Highly scalable, modular architecture with FDR Infiniband and PCIe Gen 3.0 support

Request a Quote

Future Proof System With Massive Scalability For Large Scale Deployments

The ClusterMax® Apex Supercomputer Cluster is ideal for large-scale deployments that integrate the latest CPU-based computing technologies and require extreme performance and high-density computing. Based on the 2nd Generation Intel® Xeon® Processor Scalable Family, the ClusterMax™ Apex features up to 4,032 Intel Xeon® processor cores per 42U standard rack cabinet, doubling the density compared with traditional rack mounted servers.

 

ClusterMax® Apex Supercomputer Cluster offers customers performance scalability, industry leading density and maximum efficiency at scale for applications ranging from in-memory databases, to a diverse set of data and compute-intensive HPC applications.

 

  • Delivers up to 92,160 Tensor cores, 737,280 CUDA Cores, 1,180+ Teraflops DP, 2,361+ Teraflops SP, and 18,720+ Teraflops Tensor performance per 42U cluster
  • Up to 4,608GB dedicated HBM2 GPU memory
  • Supports EDR/HDR InfiniBand fabric & real time InfiniBand diagnostics
  • Cluster management and GPU monitoring software, including GPU temperature monitoring, fan speed, and power, providing exclusive access to GPUs in a cluster

 

Applications

  • Artificial intelligence / deep learning / machine learning
  • Bio-chemical / Biotechnology / Life Sciences
  • Cloud computing
  • Computational grid endpoint
  • Computer-aided engineering (CAE)
  • Computational fluid dynamics (CFD)
  • Data mining and stream processing
  • Electronic design automation (EDA)
  • Financial market modeling
  • HPC applications (eg. Nastran, Ansys, LS-Dyna)
  • Petro-clusters / oil & gas
  • Server consolidation
  • Scientific research
  • Simulations
  • Web hosting

Cluster Specifications

  • Up to 72 nodes with 28-core 2nd Generation Intel® Xeon® Processor Scalable Family and 4,032 processor cores
  • Up to 108TB DDR4 2933/2666/2400/2133 MHz system memory
  • Up to 2,592TB hot swap storage capacity per 42U cluster
  • Great serviceability with hot-pluggable processor nodes and hot-pluggable PSU
  • High speed interconnect with FDR/EDR Infiniband, Intel® Omni-Path Architecture, Fibre channel, and Ethernet (Gigabit, 10GbE/40GbE, 25GbE, 100GbE) options
  • Dedicated on-board management port, providing a flexible and secure management environment
  • IPMI 2.0 with KVM over LAN and Virtual Media over LAN

Cluster Features

  • Delivers high availability, scalability, flexibility and power efficiency in a dense cluster architecture 
Improves RAS with hot-swappable and redundant fans, and hard disk drives
  • Addresses today’s natural business growth of mid to large-sized HPC and high-density computing
  • Lowest TCO for support, maintenance and upgradeability
  • Ideal for mid to large data centers scaling up to 1,000 nodes

Complete Cluster Assembly and Set Up Services

  • Fully integrated and pre-packaged turnkey HPC solution, including HPC professional services and support, expert installation and setup of rack-optimized cluster nodes, cabling, rails, and other peripherals
  • Configuration of cluster nodes and the network Installation of applications and client computers to offer a comprehensive solution for your IT needs
  • Rapid deployment
  • Server management options include Standards-based IPMI or AMAX remote server management
  • Seamless standard and custom application integration and cluster installation Cluster management options include a choice of open source software solutions Firmware upgrades & BIOS modification
  • Supports a variety of UPS and PDU configuration and interconnect options, including Infiniband (FDR, EDR), Fibre channel, and Ethernet (Gigabit, 10GbE, 40GbE, 25GbE, 100GbE)

Clustered File Storage (from Terabyte to Petabyte)

  • Hardware design and software stack
  • Lustre / Open source file system (Redundancy across system nodes)

Rack Level Verification

  • Performance and Benchmark Testing (HPL)
  • ATA rack level stress test
  • Rack Level Serviceability
  • Ease of Deployment Review
  • MPI jobs over IB for HPC
  • GPU stress test using CUDA
  • Cluster management

Large Scale Rack Deployment Review

  • Scalability Process
  • Rack to Rack Connectivity
  • Multi-Cluster Testing
  • Software/Application Load
  • Cluster management

Optional Cluster System Software Installed

  • Microsoft Windows Server 2019
  • Bright Computing Cluster Manager
  • SuSE / Red Hat Enterprise Linux
  • C-based software development tools, CUDA 10.x Toolkit and SDK, and various libraries for CPU GPU clusters

ClusterMax® Apex Intel Platform Standard Configuration

Total Node Count

Up to 72 nodes per 42U cluster

Processor

2x 28-core 2nd Generation Intel® Xeon® Processor Scalable Family processor series per node

Memory (per node)

16x DDR4 DIMM slots, up to 3TB DDR4 2933/2666/2400/2133 MHz ECC registered DIMMs per node

Max Cores/Threads

4,032 cores/8,064 threads

Networking

  • Dual 10GbE ports per node
  • 4X EDR InfiniBand (1 per node)
  • Supports Intel Xeon processor Scalable Family
  • Integrated fabric connectors―One 100Gb/s port per processor

Network Port Latency

1us

Management

Intel RMM4 with media redirection

Enclosure Specifications

Form Factor

2U

Systems per enclosure

4

Enclosures per rack

18

Airflow

Front to back

Power Supplies

Dual Redundant

Hard Disk Drives

12x 3.5″ or 24x 2.5″ SAS/SATA/SSD Drives

Max Disk Capacity

Up tp 16TB per drive, up to 3,456TB hot swap storage capacity per 42U cluster

Network Configuration

Switch

36 EDR/HDR InfiniBand ports

Form Factor

1U

Switching Capacity

4Tb/s

Interconnect Topology

Fat Tree

Management

  • RS232 over DB9
  • Dual 10/100/1000G Ethernet ports

Port to Port Latency

165ns

Power Supplies

Dual Redundant

Software Options

System Management

Bright Computing Bright Cluster Manager

Operating Systems

  • Red Hat Enterprise Linux (RHEL)
  • SuSE Enterprise Linux
  • CentOS
  • Ubuntu
  • Microsoft Windows Server 2019

Virtualization

  • Microsoft Windows Hyper-V
  • Citrix XenServer
  • OpenStack
  • VMware
  • ESXi/vSphere

Optimized for Turnkey Solutions

Enable powerful design, training, and visualization with built-in software tools including TensorFlow, Caffe, Torch, Theano, BIDMach cuDNN, NVIDIA CUDA Toolkit and NVIDIA DIGITS.