Today the Ethernet Alliance unveiled its 2016 Ethernet Roadmap at OFC 2016. The roadmap highlights Ethernet’s breadth of speeds, current and next-generation modules and interfaces, PoE, and innovations like the OIF’s FlexEthernet, and offers an overview of existing and future modules including QSFP-DD, microQSFP, and OBO; interfaces; and nomenclature at speeds from 10 Mb/s to 400GbE.
Zaikun Xu from the Università della Svizzera Italiana presented this talk at the Switzerland HPC Conference. “In the past decade, deep learning as a life-changing technology, has gained a huge success on various tasks, including image recognition, speech recognition, machine translation, etc. Pioneered by several research groups, Deep learning is a renaissance of neural network in the Big data era.”
Over at Enterprise Storage Forum, Henry Newman looks at why we should focus on how much work gets done rather than specifications as disk drives and SSDs get faster and faster. This is not a new rant for Henry, and in fact the importance of workflow over bandwidth or IOPs is the main theme at this year’s Mass Storage Systems and Technology Conference (MSST) coming up in May.
Even though it’s a new generation fabric, Intel OPA is still backwards compatible with the many applications in the HPC community that were written using the OpenFabrics Alliance* software stack for InfiniBand. So, existing InfiniBand users will be able to run their codes that are based on the OpenFabrics Enterprise Distribution (OFED) software on Intel OPA. Additionally, Intel has open sourced the key software elements of their fabric to allow integration of Intel OPA into the OFED stack, which several Linux* distributions include in their packages.
IBTA’s world-class compliance and interoperability program ensures the dependability of the evolving InfiniBand specification, which in turn broadens industry adoption and user confidence,” said Rupert Dance, co-chair of the IBTA Compliance and Interoperability Working Group (CIWG). “With the continued support of our members and partners, the IBTA is able to offer the industry invaluable resources to help guide critical decision making during deployment of InfiniBand or RoCE solutions.”
Today Italy’s A3Cube announced the F-730 Family of EXA-Converged parallel systems built on Dell servers and achieving sub-microsecond latency through bare metal data access. “A3Cube’s EXA-Converged infrastructure represents the next step in the evolution of converged systems”, said Emilio Billi, A3Cube’s CTO, “while keeping and improving on the scalability and resilience of Hyper-Converged infrastructure. It is engineered to converge all system resources and provide parallel data access and inter node communication at the bare metal level, eliminating the need for, and the limits of, traditional Hyper-converged systems. The system can efficiently use all the fastest storage devices currently on the market or planned to come to market, and puts all existing solutions in the rear view mirror.”
Today Cadence announced a collaboration with Mellanox Technologies to demonstrate multi-lane interoperability between Mellanox’s physical interface (PHY) IP for PCIe 4.0 technology and Cadence’s 16Gbps multi-link and multi-protocol PHY IP implemented in TSMC’s 16nm FinFET Plus (16FF+) process. Customers seeking to develop and deploy next-generation green data centers can now use a silicon-proven IP solution from Cadence for immediate integration and fastest market deployment. Cadence and Mellanox are scheduled to demonstrate electrical interoperability for PCIe 4.0 architecture between their respective PHY solutions at the 2016 TSMC Symposium on March 15, 2016 in Santa Clara, California.
The CSIRO national science agency in Australia has teamed up with Dell to deliver a new HPC cluster called “Pearcey.” The Pearcey cluster supports CSIRO research activities in a broad range of areas such as Bioinformatics, Fluid Dynamics and Materials Science. One CSIRO researcher benefiting from using Pearcey is Dr. Dayalan Gunasegaram, a CSIRO computational modeler who is using Pearcey for the modeling work behind the development of an improved nylon mesh for use in pelvic organ prolapse (POP) surgery, which has the potential to benefit the one in five Australian women that have surgery for the condition at some point in their lives.
Rich Graham presented this talk at the Stanford HPC Conference. “Exascale levels of computing pose many system- and application- level computational challenges. Mellanox Technologies, Inc. as a provider of end-to-end communication services is progressing the foundation of the InfiniBand architecture to meet the exascale challenges. This presentation will focus on recent technology improvements which significantly improve InfiniBand’s scalability, performance, and ease of use.”