- Industrial Applications of High-Performance Computing
- Featured channels
- Open Edge and HPC Initiative
Optimized AcoMod was successfully ported on Xeon Phi in native execution mode. This reduced the execution time of the code from almost 11 hours to approximately 10 minutes. The code was executed using 60 cores on one Xeon Phi card. Gromacs is a molecular dynamics code used for simulation of biochemical molecules like proteins, lipids and nucleic acids that have a lot of complicated bonded interactions. It simulates the Newtonian equations of motion for systems with hundreds to millions of particles.
This code is an open source code primarily written in C and parallelized using hybrid OpenMP and MPI models, with the compute intensive parts written in Intel intrinsic instructions.
- Persisting speech difficulties in children?
- Head of Solution Architecture - AWS– High Performance Computing - Job ID: | ingarethticyc.tk!
- The ten faces of innovation: IDEO’s strategies for beating the devil’s advocate and driving creativity throughout your organization?
A version of Gromacs in which the hotspot was coded using bit Intel intrinsic instructions for Xeon Phi was obtained and integrated into existing version of Gromacs. The code was then entirely recompiled as the input data generation steps contained Gromacs tools, which are version specific. Benchmarking of the compiled version of code was done both for host and Xeon Phi in native execution mode using Gromacs standard benchmarking input. The model employs advanced physical parameterizations, which facilitate modeling the atmospheric processes from global to mesoscale with spacing down up to metres.
Better and faster weather forecasting requires huge amount of computational resources. Recent computational accelerators like Intel Xeon Phi and Nvidia GPGPU, which have higher power efficiency, provide a better platform for weather forecasting applications. Studies were done to evaluate the performance of high resolution Nested 12 km and 4 km and Single 3 km domain WRF model configurations in host, native and symmetric execution modes.
Scalability studies using varying nodes and KNC threads were carried out.
Industrial Applications of High-Performance Computing
The execution environment has been optimized for host and symmetric mode execution and the performance bottlenecks were identified. It has been observed that single domain configurations were better suitable for many core based Xeon Phi accelerators. Ocean modelling is an inherently complex phenomenon within the earth system framework, which poses a challenge to the computational scientists. The computational requirements for ocean state forecasting are high due to the spectra of the scales of motion. Model simulations were conducted in native and symmetric execution modes on Xeon Phi.
Performance profiling of the code has been done to determine the bottlenecks, and possible improvements to achieve higher performance has been identified. Native compilation and benchmarking of VASP 5. The code was run in native execution mode and best results were obtained with total of MPI processes on 8 Xeon Phi cards. The details regarding some of the deployments are given below:.
SuMegha Cloud Builder is a tool to install cloud stack automatically for building private cloud. Sumegha Cloud Builder was enhanced with cloud middleware openStack support. MeghDoot is a comprehensive software suite designed and developed by C-DAC for building cloud computing environment. Key features includes service provisioning and deployment, ease of management through web services and enhanced security etc.
It also provides simplified graphical installation and configuration of cloud, exhaustive monitoring, metering, simplified management of resource and services, inclusion of security features focusing on data in transit, data at rest, multi-level authentication and authorization, high availability across all services and resources, backup and disaster recovery solutions etc.
Cloud Connect is an easy to use web interface for connecting clouds and simplifies the use of Infrastructure-as-a-Service IaaS feature of cloud. It abstracts creation of security group, management of network topology, creation of virtual machine, elastic block storage and snapshot, and automatic mounting of elastic block storage to virtual machine. Cloud Vault is an enterprise-class cloud storage solution offered as Storage-as-a-Service.
Users and organizations can use Cloud Vault to store large data efficiently, safely and cheaply. Next Generation Sequencing NGS is used to analyze and process the data produced as a result of genome sequencing.
Generally, the datasets produced are huge and require huge computation power and other resources. Analysis and processing of NGS data requires a work-flow where the results of one step need to be pipelined to the next step for further processing. Online NGS tool is a web-based pipeline for genome sequence analysis. Online NGS tool works on MPI-enabled virtual clusters to provide maximum computation using parallel approaches and provides storage for huge sequenced files.
This leverages cloud technology to provide logically centralized and physically distributed "Disaster Recovery-as-a-Service" model for service continuity of e-governance applications. A state-of-the-art computing and storage infrastructure for services and research was installed at Kolkata Centre that includes Tier II data centre with provision of 8 high density racks, 45 TB SAN storage and networking equipment including firewall, router with 10 Gbps backbone.
Activities involved provisioning the trending architectural components such as the Science gateways, Visualization gateways and Data grid solutions which have largely contributed to overall utility of the grid infrastructure worldwide. Grid operational activities include constant monitoring and management of grid components along with user support. It enables novice users to leverage power of Hadoop and its ecosystem components including Spark and accelerates the path to informed decisions. It comes with modest set of Big Data analysis tools which have been chosen for ease-of-use and computational power.
C-DAC has developed a Big Data analytical framework that uses multiple inputs of health care data for deriving metric based insights. C-DAC developed a Big Data Analytics Framework to analyze the genetic variations SNPs - Single Nucleotide Polymorphisms across varieties of rice genomes with origin from different countries and to visualize the level of genetic variations among them by applying machine learning techniques.
The data was taken from International Rice Informatics Consortium IRIC and the framework is developed using a combination of pre-processing, processing and post-processing tools. Pre-processing and Post-processing components are developed in-house using Python and processing is done using VariantSpark, a machine learning analysis framework for genomic data.
It facilitates researchers to visualize multiple simulation trajectory data in accelerated and efficient way. Its key feature is to load multiple trajectory files simultaneously so as to view them together and perform operations on them. Various colour coding schemes for the structures according to the users' choice are also incorporated.
NEURON provides an easy to use interface and helps researchers understand the process of identifying causality in a gene, the relationship of cause and effect. The statistical significance of predictions has been tested using multinomial coefficients derived from randomized data sets. The tool is useful for finding the target drug ligands.
The torsion angle driven conformational search method is useful in a range of chemical design applications, including drug discovery and design of targeted chemical hosts. Pharmacogenomics studies are widely adopted in clinical practices and these help in understanding the effects of drug and its dosage based on individual's genetic makeup. C-DAC has developed Big Data platform by integrating the existing pharmacogenomics data from multiple sources.
A web application has been developed with an easy to use interface for querying this integrated database and to visualize results graphically. Visualization of Pharmacogenomics data using Big data Platform. Currently, there is a need to have advanced analytics platforms and algorithms which can analyze data faster and more efficiently.enestehe.tk
Open Edge and HPC Initiative
C-DAC implemented an algorithm within the map and reduce paradigm to calculate hydrogen bonding including water-water interactions in large trajectories. Benchmarking of the algorithm brought out a linear scalability with up to 5TB of data. Toggle navigation Home. High Performance Computing. For NSM human resource development, short-term, medium-term and formal-education programs were conducted. Development and Utilization of Bioinformatics Resources and Applications Facility C-DAC has established Bioinformatics Resources and Applications Facility BRAF to provide services in the area of genome analysis, molecular modeling and systems biology including maintenance of databases, software, and high-end computing resources with application software.
Hybrid Cluster Monitoring Tool Monitoring accelerator-based hybrid clusters is imperative for early detection of any service degradation to enable immediate rectification. The Agenda Technology Alliance, a consortium focused on the paper industry, is working with two national laboratories LLNL and LBNL to computationally simulate how water flows through paper pulp during and after the pressing process. After water is initially pressed out of the paper pulp, some re-wetting typically occurs.
Currently, neither the paper de-watering nor the re-wetting phenomena are well understood, primarily due to the lack of sufficient data and reliable models describing these processes. The team is leveraging advanced simulation capabilities, experimental measurements, and paper machine data to develop an integrated, multi-physics modeling framework. Project results will be made available to the scientific community through technical presentations and publications.
Integrated steelmaking is an energy-intensive process. The steel mills feed coke, a fuel derived from coal, into blast furnaces to produce the molten iron used to make high-tonnage carbon steels.
Existing limits on modeling capabilities constrain efforts to optimize the energy efficiency of blast furnaces. Simulations are often run on desktop computers—and complex reactive flows or 3D simulations can take 30 days or more to complete.