At SuperComputing 2014 (SC14), the HPC conference in New Orleans, Louisiana, Dell has launched its Dell PowerEdge C4130 servers with Intel Xeon Phi inside. As an incredibly GPU dense and flexible HPC-focused rack server purpose-built to accelerate the most demanding HPC workloads, the PowerEdge C4130 is an Intel Xeon E5-2600v3 1U server that offers up to four GPUs/accelerators.
The Dell PowerEdge C4130 server would be able to achieve over 7.2 Teraflops on a single 1U server, with a performance/watt ratio of up to 4.17 Gigaflops per watt. This allows customers to further optimize application performance and productivity while unshackling the constraints of traditional data center space and energy use.
Incorporating innovative technology from NVIDIA and Intel, the Dell PowerEdge C4130 server is available in five uniquely designed configurations that address the flexibility necessary for varying HPC workload characteristics. This supercomputing agility allows for scientific and medical research labs, financial firms, industrial companies and organizations with analytical or data-intensive computing needs to deliver faster results and quickly adapt to changing business needs.
Higher GPU Density
The Dell PowerEdge C4130 server adds up to two Intel Xeon E5-2600v3 processors, 16 DIMMs of DDR4 memory and up to two 1.8-inch SATA SSD boot drives in just 1U of rack space. With up to four 300W Intel Xeon Phi coprocessors or NVIDIA Tesla GPU accelerators, including the new NVIDIA Tesla K80 dual-GPU accelerator, the PowerEdge C4130 would enable seamless analysis and execution of data on even the most complex and intensive computing workloads.
“A majority of HPC users are moving to higher GPU density configurations to drive discovery and innovation in seismic processing, signal and image processing, computational chemistry, and machine learning,” said Sumit Gupta, general manager of Tesla Accelerated Computing at NVIDIA. “The Dell PowerEdge C4130 server is a unique platform that packs four Tesla K40 or the new Tesla K80 dual-GPU accelerators into a compact 1U form factor, with a very flexible architecture that maximizes application performance with NVIDIA GPUDirect.”