Nvidia has introduced the new GM204 GPU based on the second-generation of the Maxwell architecture. And while the device is designed for advanced gaming graphics, it also makes for a great CUDA development platform for HPC.
The National Science Foundation is sponsoring a preliminary study to demonstrate the performance and power advantages of FPGAs over GPUs for Deep Learning in Computer Vision.
Today Eurotech announced that the company has teamed up with AppliedMicro Circuits Corporation and NVIDIA to develop a unique HPC system architecture that combines extreme density and best-in-class energy efficiency.
This week AMD announced its goal to deliver a 25x improvement in the energy efficiency of its Accelerated Processing Units (APUs) by 2020.
In this slidecast, John Gromala from HP describes the company’s new Apollo series of HPC servers. Tailor-made for the HPC market, the Apollo Series combines a modular design with innovative power distribution and air- and liquid-cooling techniques for extreme performance at rack scale, providing up to four times more performance per square foot than standard rack servers.
Heterogeneous hardware is now present in virtually all clusters. Make sure you can monitor all hardware on all installed clusters in a consistent fashion. With extra work and expertise, some open source tools can be customized for this task. There are few versatile and robust tools with a single comprehensive GUI or CLI interface that can consistently manage all popular HPC hardware and software. Any monitoring solution should not interfere with HPC workloads.
In this slidecast, Fritz Ferstl from Univa presents: Rev Up Your HPC Engine. The presentation explores the challenges for Workload Management systems in today’s datacenters with ever-increasing core counts.
“Jointly defined by a group of major computer hardware and software vendors, the OpenMP API is a portable, scalable model that gives shared memory parallel programmers a simple and flexible interface for developing parallel applications on platforms ranging from embedded systems and accelerator devices to multicore systems and shared memory systems.”
Smaller clusters often overload a single server with multiple services such as file, resource scheduling, plus monitoring/management. While this approach may work for systems with fewer than 100 nodes, these services can overload the cluster network or the single server as the cluster grows. InsideHPC Guide show a plan for scalable HPC cluster growth
“Our core product is the Xcelerit SDK, a Software Development Kit that makes it easy for domain specialists (i.e. mathematicians in banks or geophysicists in energy exploration firms) to convert their existing code to take advantage of multi-core, GPU and other hardware accelerators.”