|Surin Clusters designed with the latest intel Xeon processors. Xeon CPUs are highly suited for HPC applications with high efficiency in floating point operations, integer operations, and memory connecitvity.|
|Surin Clusters designed using the latest NVIDIA GPUs such as the Tesla V100 or P100. These GPU-based clusters provide a unique processing capability for applications and CUDA-based codes. They contain a specific ratios of CPUs to GPUs.|
|Surin Clusters based on th 8, 16, and 32 core AMD EPYC 7000 series processors. Each node provides high RAM, exceptional memory bandwidth at cost-effective performance.|
We use the fastest multi-core processors and multi-channel memory with error correction as the basis for all high-performance systems. This ensures the best performance for any software or application you or your users plan to execute.
With the use of Nvidia Tesla’s GPUs achieving higher processing power is also available.
We utilize high-capacity storage systems that have high-bandwidth and low-latency.
High-bandwidth and low-latency technologies such as the 100Gbps EDR InfiniBand and 40 Gigabit Ethernet network connections are used between nodes.
Customized and tailor made solutions based on your needs
Surin designs customized systems tailored to any budget and offers solutions based on your needs. A detailed design based on your suggestions is presented by our tem of experts. Depending on the complexity of the cluster, our design team may include HPC programmers, hardware designers, technical sales, or HPC system management specialists.
All Surin products and systems are available to be used in an HPC cluster. Each server model or workstation that matches your needs can be installed in a customized cluster. A number of silent workstations can also be used as a small computational clusters at you office.
The software and network configuration of your cluster may be specified by your experts or recommended by our experienced professionals.
The solution we present will consist of all the installation specifications, including power consumption and cooling system specs, which is required at customer’s premises for the safe operation the cluster unit. Customers can coordinate with one of our project managers to make sure their computing equipment is installed and delivered without difficulty at the customer’s premises.
Delivered Ready to Run
Surin clusters are assembled, confiured and fully tested before being sent to your site.
- Each cluster is configured with your chosen Linux version. Options include Red Hat Enterprise Linux (RHEL), CentOS, Scientific Linux, Ubuntu, SuSE Linux Enterprise Server (SLES), openSUSE, Debian and Gentoo. Microsoft Windows and Windows Server can also be installed.
- Cluster management and monitoring tools are installed, configured and tested. Options include:
- Surin Linux Cluster Management System
- HPC Windows Cluster Management System
- Open-source or commercially-supported workload manager, including SLURM, Torque, MOAB, Univa Grid Engine, or PBS Pro
- Cluster libraries and tools (such as BLAS, FFTW, MPI, Intel MKL, NVIDIA CUDA) are compiled, custom-integrated and tested.
- User-level applications and libraries are available pre-installed.
- Intel, Portland Group (PGI) and GNU compilers are available pre-installed.
- Multiple options for delivery and installation are availible depending on the budget and requirments:
- Rack Cabling at Company Place
- Pre-installed server rack rails
- Installation of ready to use clusters
- Onsite installation and training services
Technicians and sales staff will always make sure that your entire experience with Surin is prompt, creative and professional.
Phone support is available throught your cluster’s lifetime by Surin’s experienced technicians. After the end of the initial warranty period, hardware warranties are provided annually. Out-of-warranty repairs are availible on a time and matrials basis.