Supermicro, a Seagate Cloud Builder Alliance partner, has debuted new 4U 8x GPU SuperServers. The appliances are optimized for the NVIDIA Tesla K40 GPU accelerator and will provide top-flight performance for enterprises that rely on compute-intensive applications.
Supermicro made the announcement at the recent Supercomputing 2013 conference In Denver, Colo. Alongside the 4U SuperServers, enterprises will be able to benefit from Supermicro’s line of 1U, 2U and 3U SuperServers, SuperWorkstations, SuperBlade and FatTwin appliances, all of which also support the new Tesla accelerator.
“Supermicro’s HPC servers and solutions deliver the performance, scalability and reliability needed to answer the most complex challenges of our time,” stated Supermicro CEO and president Charles Liang. “Our extensive supercomputing solutions range from 5x GPU workstations to 6x GPUs in 1U/2U, 8x GPUs in 4U and 30x GPUs in 7U blade servers. Our GPU platforms are unrivaled in the industry and provide exactly the best configurations optimized for any scientific, engineering or big data analytics application.”
More specifically, Supermicro’s hardware will be beneficial to modeling, scientific research, supercomputer applications and other simulations. With the ability to house 20 GPUs in one system, Supermicro servers will provide the high-density parallel processing power needed to support these intensive use cases. The 4U 8x GPU SuperServer splits the CPU and GPU cooling zones into separate levels to improve performance and stability, supporting up to four redundant Platinum Level 1600W power supplies.
The Tesla K40 GPU is the highest-performing accelerator ever built. It is 10 times faster than the current top CPUs and 40 percent faster than its predecessor, the Tesla K20X GPU. Its power will enable a wide range of new applications that can process massive amounts of data kept in cloud storage, empowering researchers and engineers to make new discoveries.