In this video, John Carrier from Cray presents: Update on Lustre and OpenSFS. Recorded at the Open Fabrics Workshop on March 27, 2012 in Monterey, CA.
Video: Update on Lustre and OpenSFS
Video: Unified MPI and PGAS Design with RDMA to Support Hybrid Programming for Exascale
In this video, D.K. Panda from OSC presents: Unified MPI and PGAS (UPC, OpenShmem, etc.) Design with RDMA to Support Hybrid Programming Environment for Exascale systems. Recorded at the Open Fabrics Workshop on March 26, 2012 in Monterey, CA.
Video: Open Fabrics Updates and Directions
In this video, Tom Stachura from Intel presents: Open Fabrics Technology Advisory Council: Updates and Direction. Recorded at the Open Fabrics Workshop on March 26, 2012 in Monterey, CA.
Video: RSockets at the Open Fabrics Workshop
In this video, Sean Hefty from Intel presents: RSockets. Recorded at the Open Fabrics Workshop on March 26, 2012 in Monterey, CA.
Video: Josh Simons from VMware on Virtualization for HPC and Big Data
In this video, William Wallace from insideHPC interviews VMware’s Josh Simons on the topics of virtualization and RDMA, high performance computing, and Big Data. Recorded at the Open Fabrics Workshop on March 26, 2012 in Monterey, CA.
Josh has also posted some interesting notes on his presentation on RDMA at the Workshop:
In my introduction, I showed a few graphs from our upcoming paper, RDMA Performance in Virtual Machines using QDR InfiniBand on VMware vSphere 5 that showed we can use passthrough mode (VM DirectPath I/O) to deliver 1.75us half ping-pong latencies for Send and 3us for RDMA Read using polling completions and 7.6us using interrpt completions for RDMA Read.
Video: Portals 4 Exascale
In this video, Ron Brightwell from Sandia National Labs and Bob Pearson from System Fabric Works present: Portals 4 Exascale. Recorded at the Open Fabrics Workshop on March 26, 2012 in Monterey, CA.
Video: OFED Requirements
In this video, Ira Weiny from LLNL presents: OFED Requirements. Recorded at the Open Fabrics Workshop on March 26, 2012 in Monterey, CA.
Open Fabrics Workshop to Focus on HPC, Big Data, and the Cloud
This week OFA announced that the 2012 OpenFabrics Alliance User and Developer Workshop will focus on optimization and expansion in big data and cloud computing. The workshop will take place March 25-28 in Monterey.
The OFA User and Developer Workshop allows OFED users, from CIO’s, senior data center architects to mainstream network administrators, to gather in one room to share their experiences using OFED and ask how RDMA and OFED can help them achieve their business goals,” said Chris Calandro, chair of the OpenFabrics Alliance User and Developer Workshop. “This year, we’ve designed the agenda to allow for more free-flowing discussions with panels led by leading architects, developers, technologists and end users from various vertical and horizontal markets to speak openly about their experiences and lessons learned using RDMA and OFED.”
The OpenFabrics Alliance is a community that develops and supports the OFED open sourced networking software stack for HPC clusters and mainstream server cloud data centers. Companies and organizations presenting at the workshop include AMD, Chelsio, Cray, EMC, Fusion I/O, HP, Intel, Lawrence Livermore National Laboratory (LLNL), Mellanox, Microsoft, Myricom, NVIDIA, NASA, Ohio State University, Oak Ridge National Laboratory (ORNL), QLogic, Sandia, Software Forge, OCZ and VMware. Register now.
Video: Texas Instruments Multicore DSPs for HPC
In this video, Arnon Friedmann from Texas Instruments provides an overview on the company’s new Multicore DSPs for HPC.
Texas Instruments is offering HPC developers the “industry’s highest performing multicore solutions, at the lowest power levels, based on its C66x DSP generation. With the highest performing floating point DSP core at 16 GFLOPs/W, TI’s C66x multicore DSPs are changing the way HPC developers meet requirements for performance, power efficiency and ease of use.
Today TI announced that the company is also providing free optimized libraries for HPC developers, with support for standard programming languages such as C and OpenMP. With these tools, the company says that developers can easily migrate their applications to take advantage of the power savings and performance.
Open Fabrics Workshop Set for March 25, 2012
The OpenFabrics Alliance has announced dates for the 2012 OFA International Workshop: March 25-28, 2012 in Monterey, CA. The multi-day event will be focused on the strategic opportunities for the OFA and development efforts around OFED, and will include an assessment of the state of the Alliance, industry expert opinions and direction and breakout sessions from leading solution providers and OEMs discussing customer requirements and deployments.
At the 2011 OFA Workshop, it was evident that there are tremendous growth opportunities available for OFED – and OFA member companies ready to tackle the development efforts,” said Addison Snell, CEO of Intersect360 Research. “We look forward to reconnecting with everyone at the 2012 event. Our research studies continue to show a reliance on mixed interconnect environments that blend standards with performance, which matches well with the capabilities OFED is targeting.” OFED is being utilized in the majority (>60 percent) of the top 100 systems in the recently announced TOP500 list. OFED helps RDMA-based networks provide computing efficiencies as high as 96 percent and latencies as low as 1 microsecond, and supports speeds up to 56Gb/s for InfiniBand and 40Gb/s for Ethernet. OFED also reduces data center power consumption up to 50 percent.
Early bird registration for the 2012 workshop will open this fall. Read the Full Story.
Video: IB ACM InfiniBand Communication Management Assistant (for Scaling)
In this video, Intel’s Sean Hefty presents IB ACM InfiniBand Communication Management Assistant (for Scaling). Recorded at the Open Fabrics 2011 Workshop in Monterey. Slides from the Workshop are now available for download… but you’ll need to complete a free registration with OFED first.
Video: GPUs As Better MPI Citizens
In this video, Nvidia’s Dale Southard presents GPUs As Better MPI Citizens. Recorded at the Open Fabrics 2011 Workshop in Monterey. Slides from the Workshop are now available for download… but you’ll need to complete a free registration with OFED first.
Video: GPGPU Compute On AMD
In this video, AMD’s Udeepta Bordoloi presents GPGPU Compute On AMD. Recorded at the Open Fabrics 2011 Workshop in Monterey. Slides from the Workshop are now available for download… but you’ll need to complete a free registration with OFED first.
Video: Panel Discussion on OFED Open Subnet Manager
In this panel discussion moderated by LANL’s Susan Coulter, panelists discuss the Open Subnet Manager. Recorded at the Open Fabrics 2011 Workshop in Monterey. Slides from the Workshop are now available for download… but you’ll need to complete a free registration with OFED first.
Video: D.K. Panda – Can HPC Interconnects Benefit MemCached And Hadoop?
In this video, D.K. Panda and Sayantan Sur (Ohio State University) present: Can HPC Interconnects Benefit MemCached And Hadoop? Recorded at the Open Fabrics 2011 Workshop in Monterey. Slides from the Workshop are now available for download… but you’ll need to complete a free registration with OFED first.
