ClusterMax® Workgroup
Mobile Cluster Solution for Workgroup Environments
The ClusterMax® Workgroup is ideal for workgroups or small sites running HPC applications in computer-aided engineering, oil & gas, financial services, and life and material sciences, and requires no special power or cooling. The result: with the ClusterMax® Workgroup, you own the computing power you need to improve productivity, reduce design and development cycles, speed computation, improve efficiency, and reduce the cost of innovation.
The ClusterMax® Workgroup HPC Cluster was engineered as a space-saving, versatile cluster architecture that addresses today’s need for a high performance, high-density mobile cluster, enabling smaller workgroups to handle larger and more complex workloads outside of the datacenter with limited IT support. High performance, multi-processor and multi-core compute servers let you run more individual jobs concurrently and reduce time-to-solution for individual, parallelized jobs – all in a workgroup environment. The ClusterMax® Workgroup features up 32x 40-core 3rd Generation Intel® Xeon® Scalable Processors Family or AMD EPYC™ 7002 / 7003 Series Processors, or 16 A100 SXM4 GPUs per 14U standard rack, doubling the density compared with traditional rack mounted servers.
Cluster Specifications:
- Mobile/modular, and space-saving 14U HPC cluster, with support for up to:
- 32x 40-core 3rd Generation Intel® Xeon® Scalable Processors Family and 1,280 processor cores
- 32x 64-core AMD EPYC™ 7002 or 7003 Series Processors and 2,048 processor cores
- 4x AMD EPYC™ 7002 or 7003 Series Processors, 16x MVODOA A100 SXM4 GPUs, 110,592 FP32 CUDA Cores / INT32 Cores, 55,296 FP64 Cores, 6,912 Tensor Cores, 198 Teraflops of peak FP64 performance, 312 Teraflops of peak FP64 Tensor Core performance, 312 Teraflops of peak FP32 Performance
- 4x 3rd Generation Intel® Xeon® Processor Scalable Family processors, 16x MVODOA A100 SXM4 GPUs, 110,592 FP32 CUDA Cores / INT32 Cores, 55,296 FP64 Cores, 6,912 Tensor Cores, 198 Teraflops of peak FP64 performance, 312 Teraflops of peak FP64 Tensor Core
- Onboard IPMI 2.0 with KVM over IP and remote Virtual Media over LAN, enabling ease of management and control
- Highest power efficiency — 93% efficiency earth-friendly power supply
Applications:
- Climate and weather modeling, computational physics, biochemistry simulations, data analytics, satellite imaging, computational finance, deep learning, machine learning, artificial intelligence.
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 commercial and open source software solutions
- Supports a variety of UPS and PDU configuration and interconnect options, including Infiniband (EDR/HDR), Fibre channel, and Ethernet (Gigabit, 10GbE, 40GbE, 25GbE, 100GbE, 200GbE)
- Energy efficient cluster cabinets, high performance UPS and power distribution units for expert installation and setup of rack-optimized nodes, cabling, rails, and other peripherals
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
Optional Cluster System Software Installed:
- Microsoft Windows Server 2019
- Bright Computing Cluster Manager
- SuSE / Red Hat Enterprise Linux,
- C-based software development tools, CUDA Toolkit and SDK, and various libraries for CPU GPU clusters
- Deep learning software
| Model # | ClusterMax® Workgroup – I1 |
ClusterMax® Workgroup – I2 |
ClusterMax® Workgroup – A1 |
ClusterMax® Workgroup – A2 |
| Number of Nodes | 7x 1U dual socket 3rd Generation Intel® Xeon® Scalable Processor nodes | 16x 2U 4-node dual socket 3rd Generation Intel® Xeon® Scalable Processor nodes | 7x 1U dual socket AMD EPYC™ 7002 or 7003 Series Processors nodes | 16x 2U 4-node dual socket AMD EPYC™ 7002 or 7003 Series Processors nodes |
| Compute Node CPU Support | 14x processors | 32x processors | 14x processors | 32x processors |
| Compute Node Memory Support | 16 | 32 | 48 | 72 |
| GPU Memory Capacity per Rack (40GB per GPU) | 16 DIMM slots per node | 16 DIMM slots per node | 16 DIMM slots per node | 16 DIMM slots per node |
| Compute Node Storage | 4x hot-swap 3.5” HDD or 8x hot-swap 2.5” HDD/SSD per node | 12x hot-swap 3.5” HDD or 24x 2.5” HDD/SSD drive bays per enclosure | 12x hot-swap 2.5″ U.2 NVMe drive bays per node | 24x 2.5” SAS/SATA/SSD/ drives per enclosure |
| Storage Node | One 2U storage node with dual socket 3rd Generation Intel® Xeon® Scalable Processor Family, 12x hot-swap 3.5” HDD or 24x hot-swap 2.5” HDD/SSD | One 2U storage node with dual socket 3rd Generation Intel® Xeon® Scalable Processor Family, 12x hot-swap 3.5” HDD or 24x hot-swap 2.5” HDD/SSD | One 1U storage node with dual socket AMD EPYC™ 7002 or 7003 Series Processors, 12x hot-swap 2.5″ U.2 NVMe drive bays | One 1U storage node with dual socket AMD EPYC™ 7002 or 7003 Series Processors, 12x hot-swap 2.5″ U.2 NVMe drive bays |
| Storage Node CPU Support | Dual socket 3rd Generation Intel® Xeon® Processors | Dual socket 3rd Generation Intel® Xeon® Processors | Dual socket AMD EPYC™ 7002 or 7003 Series Processors | Dual socket AMD EPYC™ 7002 or 7003 Series Processors |
| Storage Node Memory | 16 DIMM slots per node | 16 DIMM slots per node | 16 DIMM slots per node | 16 DIMM slots per node |
| Rack Height | 14U 19″ rack | 14U 19″ rack | 14U 19″ rack | 14U 19″ rack |
| Network Connectivity | GbE or 10GbE connectivity on each node | GbE or 10GbE connectivity on each node | Dual 10GBase-T LAN ports on each node | Flexible IO module networking options per node |
| Network Switch | 1x 16-port GbE Ethernet switch | 1x 16-port GbE Ethernet switch | 10GBase-T Stackable managed switch | 10GBase-T Stackable managed switch |
| Infiniband Switch | Optional EDR/HDR infiniband switch | Optional EDR/HDR infiniband switch | Optional EDR/HDR infiniband switch | Optional EDR/HDR infiniband switch |
| Cluster Management Software | Bright Cluster Manager software options with 1-year or 3-year support | Bright Cluster Manager software options with 1-year or 3-year support | Bright Cluster Manager software options with 1-year or 3-year support | Bright Cluster Manager software options with 1-year or 3-year support |
Software Options
Bright Cluster Manager software automates the process of building and managing modern high-performance Linux clusters, eliminating complexity and enabling flexibility.
NVMesh enables shared NVMe across any network and supports any local or distributed file system. The solution features an intelligent management layer that abstracts underlying hardware with CPU offload, creates logical volumes with redundancy, and provides centralized, intelligent management and monitoring.
QuantaStor’s unique Storage Grid architecture organizations are able to manage multiple clusters across sites as a unified storage platform that’s easily configured and maintained through the web user interface and automated via advanced CLI and REST APIs
Enabling data centers to easily transform themselves into a flexible cloud infrastructure with the performance and reliability needed to run enterprise applications.
| Model # | ClusterMax® Workgroup – GX100S | ClusterMax® Workgroup – GA100S |
| Rack Height | 14U | 14U |
| # of 4U 8x A100 SXM4 GPU Nodes per rack | 2 | 2 |
| # of A100 SXM4 GPUs per Rack (8x GPU per Node) | 16 | 16 |
| GPU Memory Capacity per Rack (40GB per GPU) | 640GB | 640GB |
| GPU Memory Capacity per Rack (80GB per GPU) | 1,280GB | 1,280GB |
| GPU Node Processor Support | 2x 3rd Generation Intel® Xeon® Processor per node | 2x AMD EPYC™ 7002 or 7003 Series Processors per node |
| # of Processors per Rack ( 2 Processors per node) | 4 | 4 |
| Maximum # of CPU Cores per Rack (64 cores per Processor) | 320 Cores | 320 Cores |
| Maximum Compute Node Memory Capacity per Rack (8TB per system) | 32TB | 32TB |
| # of FP32 CUDA Cores per Rack (6,912 cores per GPU) | 110,592 Cores | 110,592 Cores |
| # of FP64 Cores per Rack (3,456 cores per GPU) | 55,296 Cores | 55,296 Cores |
| # of INT32 Cores per Rack (6,912 cores per GPU) | 110,592 Cores | 110,592 Cores |
| Maximum Compute Node Memory Capacity per Rack (8TB per system) | 32TB | 32TB |
| # of Tensor Cores per Rack (432 cores per GPU) | 6,912 Cores | 6,912 Cores |
| Peak FP64 Performance per Rack (9.7 TF per GPU) | 155 TFLOPS | 155 TFLOPS |
| Peak FP64 Tensor Core Performance per Rack (19.5 TF per GPU) | 312 TFLOPS | 312 TFLOPS |
| Peak FP32 Performance per Rack (19.5 TF per GPU) | 312TFLOPS | 312TFLOPS |
| Tensor Float 32 (TF32) Performance per Rack (156 TF per GPU) | 2,496 TFLOPS | 2,496 TFLOPS |
| Tensor Float 32 (TF32) Performance per Rack, with Sparsity (312 TF per GPU) | 4,992 TFLOPS | 4,992 TFLOPS |
| Peak BFLOAT16 / FP16 tensor Core Performance per Rack (312 TF per GPU) | 4,992 TFLOPS | 4,992 TFLOPS |
| Peak BFLOAT16 / FP16 tensor Core Performance per Rack, with Sparsity (624 TF per GPU) | 9,984 TFLOPS | 9,984 TFLOPS |
| Peak INT8 tensor Core Performance per Rack (624 TOPs per GPU) | 9,984 TOPs | 9,984 TOPs |
| Peak INT8 tensor Core Performance per Rack, with Sparsity (1,248 TOPs per GPU) | 19,968 TOPs | 19,968 TOPs |
| Peak INT4 tensor Core Performance per Rack (1,248 TOPS per GPU) | 19,968 TOPs | 19,968 TOPs |
| Peak INT4 tensor Core Performance per Rack, with Sparsity (2,496 TOPs per GPU) | 39,936 TOPs | 39,936 TOPs |
| GPU Nodes Interconnectivity | 10GbE | 10GbE |
| GPU Node Storage | 6x U.2 NVMe bays & 2x M.2 NVMe bays | 6x U.2 NVMe bays & 2x M.2 NVMe bays |
| Network Switch | 1x 24-port 10GbE Gigabit Ethernet | 1x 24-port 10GbE Gigabit Ethernet |
| Cluster Management Software | Optional Bright Cluster Manager software | Optional Bright Cluster Manager software |
Software Options
Bright Cluster Manager software automates the process of building and managing modern high-performance Linux clusters, eliminating complexity and enabling flexibility.
NVMesh enables shared NVMe across any network and supports any local or distributed file system. The solution features an intelligent management layer that abstracts underlying hardware with CPU offload, creates logical volumes with redundancy, and provides centralized, intelligent management and monitoring.
QuantaStor’s unique Storage Grid architecture organizations are able to manage multiple clusters across sites as a unified storage platform that’s easily configured and maintained through the web user interface and automated via advanced CLI and REST APIs
Enabling data centers to easily transform themselves into a flexible cloud infrastructure with the performance and reliability needed to run enterprise applications.