Commenced in January 2007
Frequency: Monthly
Edition: International
Paper Count: 4

MPI Related Abstracts

4 Performance Evaluation of Soft RoCE over 1 Gigabit Ethernet

Authors: Manoj Kumar, Manju Bala, Gurkirat Kaur

Abstract:

Ethernet is the most influential and widely used technology in the world. With the growing demand of low latency and high throughput technologies like InfiniBand and RoCE, unique features viz. RDMA (Remote Direct Memory Access) have evolved. RDMA is an effective technology which is used for reducing system load and improving performance. InfiniBand is a well known technology which provides high-bandwidth and low-latency and makes optimal use of in-built features like RDMA. With the rapid evolution of InfiniBand technology and Ethernet lacking the RDMA and zero copy protocol, the Ethernet community has came out with a new enhancements that bridges the gap between InfiniBand and Ethernet. By adding the RDMA and zero copy protocol to the Ethernet a new networking technology is evolved, called RDMA over Converged Ethernet (RoCE). RoCE is a standard released by the IBTA standardization body to define RDMA protocol over Ethernet. With the emergence of lossless Ethernet, RoCE uses InfiniBand’s efficient transport to provide the platform for deploying RDMA technology in mainstream data centres over 10GigE, 40GigE and beyond. RoCE provide all of the InfiniBand benefits transport benefits and well established RDMA ecosystem combined with converged Ethernet. In this paper, we evaluate the heterogeneous Linux cluster, having multi nodes with fast interconnects i.e. gigabit Ethernet and Soft RoCE. This paper presents the heterogeneous Linux cluster configuration and evaluates its performance using Intel’s MPI Benchmarks. Our result shows that Soft RoCE is performing better than Ethernet in various performance metrics like bandwidth, latency and throughput.

Keywords: Ethernet, InfiniBand, RoCE, RDMA, MPI, Soft RoCE

Procedia PDF Downloads 258
3 The Integrated Methodological Development of Reliability, Risk and Condition-Based Maintenance in the Improvement of the Thermal Power Plant Availability

Authors: Isti Surjandari, Iwa Garniwa, Henry Pariaman, Bambang Sugiarto

Abstract:

Availability of a complex system of thermal power plant is strongly influenced by the reliability of spare parts and maintenance management policies. A reliability-centered maintenance (RCM) technique is an established method of analysis and is the main reference for maintenance planning. This method considers the consequences of failure in its implementation, but does not deal with further risk of down time that associated with failures, loss of production or high maintenance costs. Risk-based maintenance (RBM) technique provides support strategies to minimize the risks posed by the failure to obtain maintenance task considering cost effectiveness. Meanwhile, condition-based maintenance (CBM) focuses on monitoring the application of the conditions that allow the planning and scheduling of maintenance or other action should be taken to avoid the risk of failure prior to the time-based maintenance. Implementation of RCM, RBM, CBM alone or combined RCM and RBM or RCM and CBM is a maintenance technique used in thermal power plants. Implementation of these three techniques in an integrated maintenance will increase the availability of thermal power plants compared to the use of maintenance techniques individually or in combination of two techniques. This study uses the reliability, risks and conditions-based maintenance in an integrated manner to increase the availability of thermal power plants. The method generates MPI (Priority Maintenance Index) is RPN (Risk Priority Number) are multiplied by RI (Risk Index) and FDT (Failure Defense Task) which can generate the task of monitoring and assessment of conditions other than maintenance tasks. Both MPI and FDT obtained from development of functional tree, failure mode effects analysis, fault-tree analysis, and risk analysis (risk assessment and risk evaluation) were then used to develop and implement a plan and schedule maintenance, monitoring and assessment of the condition and ultimately perform availability analysis. The results of this study indicate that the reliability, risks and conditions-based maintenance methods, in an integrated manner can increase the availability of thermal power plants.

Keywords: Availability, MPI, integrated maintenance techniques, thermal power plant, FDT

Procedia PDF Downloads 649
2 Detecting the Edge of Multiple Images in Parallel

Authors: U. Dinesh Acharya, Prakash K. Aithal, Rajesh Gopakumar

Abstract:

Edge is variation of brightness in an image. Edge detection is useful in many application areas such as finding forests, rivers from a satellite image, detecting broken bone in a medical image etc. The paper discusses about finding edge of multiple aerial images in parallel .The proposed work tested on 38 images 37 colored and one monochrome image. The time taken to process N images in parallel is equivalent to time taken to process 1 image in sequential. The proposed method achieves pixel level parallelism as well as image level parallelism.

Keywords: multicore, gpu, Edge Detection, MPI, opencl

Procedia PDF Downloads 188
1 Parallelizing the Hybrid Pseudo-Spectral Time Domain/Finite Difference Time Domain Algorithms for the Large-Scale Electromagnetic Simulations Using Massage Passing Interface Library

Authors: Donggun Lee, Q-Han Park

Abstract:

Due to its coarse grid, the Pseudo-Spectral Time Domain (PSTD) method has advantages against the Finite Difference Time Domain (FDTD) method in terms of memory requirement and operation time. However, since the efficiency of parallelization is much lower than that of FDTD, PSTD is not a useful method for a large-scale electromagnetic simulation in a parallel platform. In this paper, we propose the parallelization technique of the hybrid PSTD-FDTD (HPF) method which simultaneously possesses the efficient parallelizability of FDTD and the quick speed and low memory requirement of PSTD. Parallelization cost of the HPF method is exactly the same as the parallel FDTD, but still, it occupies much less memory space and has faster operation speed than the parallel FDTD. Experiments in distributed memory systems have shown that the parallel HPF method saves up to 96% of the operation time and reduces 84% of the memory requirement. Also, by combining the OpenMP library to the MPI library, we further reduced the operation time of the parallel HPF method by 50%.

Keywords: Hybrid, parallelization, MPI, OpenMP, FDTD, PSTD

Procedia PDF Downloads 5