| Viglen launch Nehalem products... read on > |
| Viglen HPC Services — with you all the way... read on > |
| Remote HPC Management Providing Peace of Mind... read on > |
| Linux HPC Solution - Making Life easy and straightforward... read on > |
Head Nodes| Queen Mary University |
| University Hertfordshire |
| University Bedfordshire |
| all case studies |
 HPC Pricelist |
 HPC flyer |
|
• HX420Hi Head Node |
Supercharge your cluster with Tesla CUDA-based GPUs
The Viglen Tesla Preconfigured Clusters deliver supercomputing performance at a lower power, lower cost, and using much fewer systems than standard CPU-only clusters. These clusters are powered by the NVIDIA® Tesla™ S1070 Computing System, a four-teraflop 1U system that has the Tesla T10 GPUs based on the CUDA massively parallel architecture. With 4 GB of memory per GPU, support for IEEE 754 single & double floating point precision, and a fast 102 GB/sec GDDR3 interface to the memory, the S1070 1U speeds the transition to energy-efficient parallel computing and scales to solve the world's most important computing challenges more quickly and accurately. Successfully deployed at research institutes, universities, and enterprises — it powers the 29th most powerful supercomputer in the world.
Viglen offers several configurations of the Tesla preconfigured clusters and we can customise them to your needs:
| Configurations | ||
| Cluster | ||
| Sixteen Tesla S1070 Preconfigured Cluster | Eight Tesla S1070 Preconfigured Cluster | Four Tesla S1070 Preconfigured Cluster |
| GPUs | ||
| 64 Tesla T10 GPUs | 32 Tesla T10 GPUs | 16 Tesla T10 GPUs |
| CPU Servers | ||
| Eight HX420T2i Quad Motherboard, Dual-socket Quad-core Nehalem | Four HX420T2i Quad Motherboard, Dual-socket Quad-core Nehalem | Two HX420T2i Quad Motherboard, Dual-socket Quad-core Nehalem |
| Head Node | ||
| One HX420Hi with Dual-socket Quad-core Nehalem | One HX420Hi with Dual-socket Quad-core Nehalem | One HX420Hi with Dual-socket Quad-core Nehalem |
| Memory | ||
| 512 GB GPU memory. Up to 512 GB CPU memory | 256 GB GPU memory. Up to 256 GB CPU memory | 128 GB GPU memory. Up to 128 GB CPU memory |
| Storage | ||
| Up to 64 TB SATA drives | Up to 32 TB SATA drives | Up to 16 TB SATA drives |
| Network | ||
| QDR Infiniband on compute nodes GigE Switch for cluster | QDR Infiniband on compute nodes GigE Switch for cluster | QDR Infiniband on compute nodes GigE Switch for cluster |
Pay as you grow
Viglen Expertise and Leading Edge Technology
Each Preconfigured Tesla Cluster is supplied as a complete validated, tested and certified solution. Designed to Intel Cluster Ready programme specifications supplied with software specifically tuned to maximise the performance of your cluster and CUDA based parallel tasks.
The host system, the Viglen HX420T2i is based on leading edge Intel Nehalem technology, offering fast PCI Gen2.0 x16 connectivity to the S1070 GPU nodes, high efficiency redundant power, hot-pluggable motherboard technology, latest generation industry leading CPU performance, QDR Infiniband connectivity designed to minimise latency whilst offering a massive 40Gb/s throughput from each compute node.
Tightly coupling this hardware and ensuring the system operates a single manageable unit is the Cluster Management Stack, ClusterCorps Rocks+. Rocks is one of the industry’s most mature cluster management stacks (with over 10 years of development) and most widely used (with over 10,000 clusters installed worldwide). Integrated in this cluster stack is the Rocks+ CUDA roll which deploys the CUDA environment and modules to all nodes, the Rocks* Mellanox OFED stack and the Rocks+ Intel Cluster Ready roll which provides a Cluster Health and Performance Checker for the cluster.
The Viglen Preconfigured Tesla Cluster solutions offer a complete turnkey supported high performance computing solution.
Installed Software
The Viglen Tesla preconfigured cluster comes with preinstalled with:
- ClusterCorp Rocks+ Linux Cluster Management Stack
- CUDA 2.2 Toolkit and SDK (Rocks+ CUDA Roll)
- Integrated Clustering Software inc:
- Mellanox OFED Roll (QDR Infiniband)
- Intel Cluster Ready Roll (inc Cluster Health and Performance Checker)
CUDA Applications
The CUDA-based Tesla GPUs give speedups of up to 250x on applications ranging from MATLAB to computational fluid dynamics, molecular dynamics, quantum chemistry, imaging, signal processing, bio-informatics, and so on. Click here to learn more about these speedups with links to application downloads.
GPU Software Development Tools
Click here to find out more about the C-based software development tools and various libraries for GPUs.
Configuration to enter the top 500 Supercomputing list | |
| Cluster | |
| Forty two Tesla S1070 Preconfigured Cluster | |
| GPUs | |
| 336 Tesla T10 GPUs | |
| CPU Servers | |
| Twenty one HX420T2i Quad Motherboard, Dual-socket Quad-core Nehalem | |
| Head Node | |
| One HX420Hi with Dual-socket Quad-core Nehalem | |
| Memory | |
| 1344GB GPU memory. Up to 1344 GB CPU memory | |
| Storage | |
| Up to 168TB SATA drives | |
| Network | |
| QDR Infiniband on compute nodes GigE Switch for cluster | |
Please call your Account Manager for further information.
