| Products |
| |  | Clusters |
| |  | Servers |
|
|
| |  | Storage |
|
|
| |  |
Parallel Computing |
|
|
|
|
| Questions? |
|
|
|
Call 1.800.800.6328
|
|
|
|
|
|
 |
|
 |
|
| Testimonial |
"We have entrusted AMAX for over 10 years to handle all of our Enterprise IT computing systems. There is a great sense of relief knowing we can depend upon AMAX's technical staff, their recommendations related to cost pricing, support of the product when it should arise, and sense of customer satisfaction."
- Jeff Turner, Network Supervisor, Lafayette School District |
|
Feedback
|
|
| Preconfigured AMAX Tesla SimCluster | | Reducing Simulation Time from Days to Hours |
 |
 |
Scale your science with AMAX's pre-configured Tesla™ GPU SimCluster, an off-the-shelf and ready-to-deploy cluster solution with both hardware and software. Available in configurations that supports the NVIDIA’s Kepler (K10/K20/K20X) GPU Computing Accelerators, it delivers the easiest way to buy, use and manage GPU-based research cluster. With AMAX's Tesla GPU SimCluster, academic research departments can take GPU-accelerated apps from workstation to multi-node cluster, speeding research and discovery.
| Key Features | K20 GPU Clusters | K10 GPU Clusters | Applications | | AMAX Pre-Configured ClusterMax SimCluster Key Features: Enabling Breakthrough Science for Researchers Leading researchers in the field of life sciences, engineering, physics, and mathematics are leveraging GPUs to solve their most challenging problems. With accelerations of up to 10x, researchers are now able to run computer simulations which were previously impossible. For example, GPUs enable large, realistic biological systems to be simulated in minutes or hours rather than days and weeks. Accelerate Applications Accelerate your application with directives and GPUs. Directives compiler offers a fast and easy approach to GPU computing and can deliver 2x application performance increase in 4 weeks. Guaranteed. Scale Your Science on a GPU Cluster - Makes more complex and realistic simulation possible with massively parallel processing power of GPUs o Reduce simulation time from days to hours
- Computing cluster with GPUs as parallel co-processors designed to out-perform those built with CPUs alone
- Takes your GPU-accelerated research apps from workstation to multi-node HPC cluster
Easiest Way to Buy, Use, and Manage a GPU-based Research Cluster - Turnkey, preconfigured, requiring minimum set up
- Configured with optional software development environment and cluster
| Pre-configured AMAX NVIDIA® Tesla™ Kepler K20/K20X-Based SimCluster Specifications: | Model # | ClusterMax™ SimCluster-K20.14U | ClusterMax™ SimCluster-K20.24U | ClusterMax™ SimCluster-K20.42U | | # of GPU Nodes | 4x 1U | 8x 1U | 16x 1U | | Included CPUs / GPU Node | 2x Intel Xeon E5-2650 2.0GHz, with 8 cores & 16 threads | 2x Intel Xeon E5-2650 2.0GHz, with 8 cores & 16 threads | 2x Intel Xeon E5-2650 2.0GHz, with 8 cores & 16 threads | | Included Memory / GPU Node | 32GB DDR3 1333MHz (upgradeable to 256GB) | 64GB DDR3 1333MHz (upgradeable to 256GB) | 64GB DDR3 1333MHz (upgradeable to 256GB) | | Included GPU / Node | 3x Tesla K20/K20X, with 12 GPUs per cluster | 3x Tesla K20/K20X, with 24 GPUs per cluster | 3x Tesla K20/K20X, with 48 GPUs per cluster | | Included GPU Memory | 72GB | 144GB | 288GB | | # of GPU Cores Included | 32,256 | 64,512 | 129,024 | | SP Performance Included | 47+ Teraflops | 94+ Teraflops | 189+ Teraflops | | DP Performance Included | 15+ Teraflops | 31+ Teraflops | 63+ Teraflops | | Interconnectivity on GPU Nodes | InfiniBand | InfiniBand | InfiniBand | | Included Storage Capacity / GPU Node | 1TB | 4TB | 4TB | | Master Node | 1x 1U | 1x 1U | 1x 1U | | CPUs / Master Node | 1x Intel Xeon E5-2620 2.0GHz (upgradeable to two processors) | 1x Intel Xeon E5-2620 2.0GHz (upgradeable to two processors) | 1x Intel Xeon E5-2620 2.0GHz (upgradeable to two processors) | | Memory / Master Node | 32GB DDR3 1333MHz (upgradeable to 256GB) | 32GB DDR3 1333MHz (upgradeable to 256GB) | 32GB DDR3 1333MHz (upgradeable to 256GB) | | Storage Capacity / Mater Node | 1TB | 1TB | 1TB | | Included Interconnectivity on Master Node | InfiniBand | InfiniBand | InfiniBand | | Rack Cabinet | 14U | 24U | 42U | | Network Switch | 1x 24-port Gigabit Ethernet 1x 18-Port InfiniBand | 1x 24-port Gigabit Ethernet 1x 18-Port InfiniBand | 1x 24-port Gigabit Ethernet 1x 18 -Port InfiniBand | | Reference # | Q160781 (Compute) + Q160775 (head) + Q159280 (cabinet) | Q160782 (Compute) + Q160775 (head) + Q159281 (cabinet) | Q160783 (Compute) + Q160775 (head) + Q159259 (cabinet) | Pre-configured AMAX NVIDIA® Tesla™ Kepler K10-Based SimCluster Specifications: | Model # | ClusterMax™ SimCluster-K10.14U | ClusterMax™ SimCluster-K10.24U | ClusterMax™ SimCluster-K10.42U | | # of GPU Nodes | 4x 1U | 8x 1U | 16x 1U | | Included CPUs / GPU Node | 2x Intel Xeon E5-2650 2.0GHz, with 8 cores & 16 threads | 2x Intel Xeon E5-2650 2.0GHz, with 8 cores & 16 threads | 2x Intel Xeon E5-2650 2.0GHz, with 8 cores & 16 threads | | Included Memory / GPU Node | 32GB DDR3 1333MHz (upgradeable to 256GB) | 64GB DDR3 1333MHz (upgradeable to 256GB) | 64GB DDR3 1333MHz (upgradeable to 256GB) | | Included GPU / Node | 3x Tesla Kepler K10, with 12 GPUs per cluster | 3x Tesla Kepler K10, with 24 GPUs per cluster | 3x Tesla Kepler K10, with 48 GPUs per cluster | | Included GPU Memory | 96GB | 192GB | 384GB | | # of GPU Cores Included | 15,360 | 30,720 | 61,440 | | SP Performance Included | 55 Teraflops | 110 Teraflops | 220 Teraflops | | DP Performance Included | 2.64 Teraflops | 5.28 Teraflops | 10.56 Teraflops | | Interconnectivity on GPU Nodes | InfiniBand | InfiniBand | InfiniBand | | Included Storage Capacity / GPU Node | 1TB | 4TB | 4TB | | Master Node | 1x 1U | 1x 1U | 1x 1U | | CPUs / Master Node | 1x Intel Xeon E5-2620 2.0GHz (upgradeable to two processors) | 1x Intel Xeon E5-2620 2.0GHz (upgradeable to two processors) | 1x Intel Xeon E5-2620 2.0GHz (upgradeable to two processors) | | Memory / Master Node | 32GB DDR3 1333MHz (upgradeable to 256GB) | 32GB DDR3 1333MHz (upgradeable to 256GB) | 32GB DDR3 1333MHz (upgradeable to 256GB) | | Storage Capacity / Mater Node | 1TB | 1TB | 1TB | | Included Interconnectivity on Master Node | InfiniBand | InfiniBand | InfiniBand | | Rack Cabinet | 14U | 24U | 42U | | Network Switch | 1x 24-port Gigabit Ethernet 1x 18-Port InfiniBand | 1x 24-port Gigabit Ethernet 1x 18-Port InfiniBand | 1x 24-port Gigabit Ethernet 1x 18 -Port InfiniBand | | Reference # | Q160776 (Compute) + Q160775 (head) + Q160777 (cabinet) | Q160778 (Compute) + Q160775 (head) + Q159281 (cabinet) | Q160779 (Compute) + Q160775 (head) + Q160780 (cabinet) | Many popular scientific and engineering codes are accelerated on the GPU today, including: | Life Sciences | Engineering | Math / Physics | | AMBER | ANSYS Mechanical | Chroma | | GROMACS | CST MICROWAVE STUDIO | Mathematica | | LAMMPS | FluiDyna OpenFOAM | MATLAB | | NAMD | SIMULIA Abaqus | MILC | |
|
Data Sheet Download
|
|
 |
copyright |
|