High Performance Computing (HPC): How it Works…
Our High Performance Computing Clusters (or HPCC for short) are designed to enable high performance parallel computing and consist of one or more master nodes (the Master) and one or more compute nodes (the Slaves) interconnected by a private ultra high performance network infrastructure (Infiniband, 1GBe or 10GBe Fibre, or 1-2GBe Copper). Each HPCC (or stacks as they are more commonly known) are built using our high performance RS series of 1U rack, or Deskside (desktop) compute clusters.
The ‘master’ node is designed to provide the stack with its main gateway to external resources and supports the Network File System (NFS). The compute ‘slave’ nodes are the cluster workhorses and are designed to perform parallel computing tasks (deployed by the parallel software applications). Typically, access and management of slave nodes are provided via remote interfaces, such as their built in IPM Interface or serial port connections through the ‘master’ node.
Our HPC solutions are built using the very latest range of 64Bit multi-core processors and interleaved memory. With support for up to 64x processing threads, and 512GB of RAM per compute node we offer the highest levels of performance.
Available with a host of CPU GHz Speeds and TDP Footprints, plus offering full support for the latest in GPGPU technologies (NVIDIA Tesla GPU Computing with CUDA, OpenCL etc).
High Performance Computing Solutions with Great Return on Investment (ROI)
High Performance Computing Solutions with Streamlined Deployment and Management
We then streamline the implementation process by validating and pre-configuring each cluster node to meet a wide variety of application needs (CAD/CAM/CAE Analysis, 3D Image Rendering, Oil/Gas, Scientific Calculations etc). Our unique HPC solutions deliver super computing performance to business’s of any size, whilst reducing cost and complexity of ownership.