Imagine what might occur if our top researchers and scientists were lacking the essential tools to help pave the way for new discoveries. Think of the potential missed opportunities and technological advancements that would be overlooked. While these brilliant minds are the birthplace of big ideas, the technology that researchers, engineers and scientists use truly help bring innovative ideas to life.
In order to serve customers such as scientists and engineers who are looking for ways to expedite the process of research and getting results, Dell today introduced the PowerEdge C410x. Back in March, we introduced the PowerEdge C family of servers, which were designed with inspiration from our Data Center Solutions (DCS) customers. The C410x expands this family and adds non-server solutions to help a greater number of customers build efficient and affordable hyperscale data center infrastructures; the C410x rounds out a portfolio of solutions that enable customers to grow and thrive.
The C410x, an external PCI Express (PCIe) expansion chassis, delivers first-of-its-kind expansion ability to help power groundbreaking research. It connects host nodes to up to 16 PCIe devices to get results faster, while saving on space, weight and cost. It is available and qualified with NVIDIA Tesla M1060 or M2050 computing modules, graphics processing units (GPUs) and qualified with the PowerEdge C6100 server.
Below, Joe Sekel, systems architect for DCS, provides a demonstration of the ins and outs of the PowerEdge C410x.
The key to success for DCS has been identifying underserved market segments and innovating to meet the specific needs of customers in those segments in a way that no other vendor has. The C410x institutes more aggressive power-saving tactics while packing a mighty punch with processing speed to provide more efficiency for researching, movie rendering, and gaming environments. Because in the end, the brilliant minds that tackle these niche markets need brilliant tools.
To post a comment login or create an account.
Unrelated comments or requests for service will be unpublished. Please post your technical questions in the Support Forums or for direct assistance contact Dell Customer Service or Dell Technical Support.. All comments must adhere to the Dell Community Terms of Use.
While the name GPGPU , which stands for General-purpose computing on graphics processing units, doesn’t
In my last entry I talked about the wild and wacky world of GPGPUs and provided an overview of the PowerEdge
In the last of my GPGPU/ PowerEdge C410x trilogy I offer up a whiteboard session with the system’s
This has potential to solve many problems to a variety of HPC customers. For floating point intensive apps and customers (Oil & Gas), having several GPUs available for offload is a great solution, but there are other use cases that could benefit in a different way. The chassis could also be filled with flash memory based of even DRAM based storage cards(like Joe referenced in the YouTube video - FusionIO, STEC, etc.) and used in a variety of different ways. For high I/O patterns, large chunks of /tmp space could be dynamically allocated and then torn down when the task was complete. For extremely large memory applications, large chunks of SWAP can be dynamically added to ease the pain (this is not a perfect substitute for DRAM, but for some cases the amount of memory needed can be cost prohibitive, so this could be a cost effective alternative). I can also think of scenarios where I/O cards could be pooled, and dynamically allocated for load based optimizations. And then there is the scenario where you fill the chassis with a variety of cards and dynamically allocated from the shared pool as load demands and hand back to the pool when done...