“HBM is a new type of CPU/GPU memory (“RAM”) that vertically stacks memory chips, like floors in a skyscraper. In doing so, it shortens your information commute. Those towers connect to the CPU or GPU through an ultra-fast interconnect called the “interposer.” Several stacks of HBM are plugged into the interposer alongside a CPU or GPU, and that assembled module connects to a circuit board. Though these HBM stacks are not physically integrated with the CPU or GPU, they are so closely and quickly connected via the interposer that HBM’s characteristics are nearly indistinguishable from on-chip integrated RAM.”
Today One Stop Systems showcased its High Density Compute Accelerator (HDCA) with AMD FirePro S9170 GPUs at SC15. The HDCA, with sixteen AMD FirePro S9170 GPUs, provides up to 84 Tflops of peak single precision and 512GB of GPU memory. The HDCA is a complete appliance, cabled from one to four host servers through PCIe x16 Gen3 connections and perfect for adding compute power to any HPC application.
Lewey Anton reports on who’s jumping ship and moving on up in high performance computing. Familiar names this month include Ron Blessing, Rick Friedman, Karl Freund, and Mark Woodyard.
The AMD graphics cards are uniquely equipped with AMD Multiuser GPU technology embedded into the GPU delivering consistent and predictable performance,” said Sean Burke, AMD corporate vice president and general manager, Professional Graphics. “When these AMD GPUs are appropriately configured to the needs of an organization, end users get the same access to the GPU no matter their workload. Each user is provided with the virtualized performance to design, create and execute their workflows without any one user tying up the entire GPU.”
“The AMD FirePro S9170 server GPU can accelerate complex workloads in scientific computing, data analytics, or seismic processing, wielding an industry-leading 32GB of memory. We designed the new offering for supercomputers to achieve massive compute performance while maximizing available power budgets.”
In this slidecast, Einar Rustad from Numascale describes how the company achieved a world-record on the McCalpin STREAM benchmark using their innovative scale-out to scale-up architecture. The benchmark measures sustainable memory bandwidth and the corresponding computation rate for simple vector kernels.
Today AMD unveiled innovation in heterogeneous HPC at the Centre of New Technologies at the University of Warsaw. In a new cluster deployment called Orion, the Next Generation Sequencing Centre in Warsaw is powering bioinformatics research with 1.5 petaFLOPS of AMD FirePro S9150 server GPU performance.
In this video, researchers describes the Lattice-CSC cluster in Darmstadt, Germany. Currently the most energy-efficient supercomputer on the Green500 list, the Lattice-CSC is powered by 640 AMD FirePro S9150 GPUs to achieve 5.27 GFLOPS/Watt.
In the second article in his series on future hardware for HPC, Robert Roe from Scientific Computing World looks at alternatives to the x86, including server on a chip and OpenPower.
Dr. David Rohr from the Frankfurt Institute for Advanced Studies (FIAS) describes how he is using AMD FirePro to accelerate high energy physics.