Papers by Ibrahim Muhammad Umar
IEEE International Conference on High Performance Computing, Data, and Analytics, May 7, 2012
ABSTRACT Massively parallel platforms have been genuinely affordable over recent years. This brea... more ABSTRACT Massively parallel platforms have been genuinely affordable over recent years. This breakthrough is the key factor that permits various research work that are continuously seeking for a better run-time performance of algorithms that process ever-growing amount of data. Multiple sequence alignment has been an important-yet-complex piece of algorithm in the computational sequence analysis field that naturally required to process huge sequence data. This proposed research looks into the strategies to utilise the massively parallel platforms in the quest of pursuing a better multiple sequence alignment in terms of throughput and parallel scalability. The initial strategies as well as the expected challenges are elaborated in this early phase of research. Current achievements as well as formulated future directions are discussed at the near end of this preliminary study.
Journal of STEM Teacher Education, 2010
IEEE Transactions on Parallel and Distributed Systems, 2019
Concurrent search trees are crucial data abstractions widely used in many important systems such ... more Concurrent search trees are crucial data abstractions widely used in many important systems such as databases, file systems and data storage. Like other fundamental abstractions for energy-efficient computing, concurrent search trees should support both high concurrency and fine-grained data locality in a platform-independent manner. However, existing portable fine-grained locality-aware search trees such as ones based on the van Emde Boas layout (vEB-based trees) poorly support concurrent update operations while existing highlyconcurrent search trees such as non-blocking search trees do not consider fine-grained data locality. In this paper, we first present a novel methodology to achieve both portable fine-grained data locality and high concurrency for search trees. Based on the methodology, we devise a novel locality-aware concurrent search tree called GreenBST. To the best of our knowledge, GreenBST is the first practical search tree that achieves both portable fine-grained data locality and high concurrency. We analyze and compare GreenBST energy efficiency (in operations/Joule) and performance (in operations/second) with seven prominent concurrent search trees on a high performance computing (HPC) platform (Intel Xeon), an embedded platform (ARM), and an accelerator platform (Intel Xeon Phi) using parallel micro-benchmarks (Synchrobench). Our experimental results show that GreenBST achieves the best energy efficiency and performance on all the different platforms. GreenBST achieves up to 50% more energy efficiency and 60% higher throughput than the best competitor in the parallel benchmarks. These results confirm the viability of our new methodology to achieve both portable fine-grained data locality and high concurrency for search trees.
Lecture Notes in Computer Science, 2016
Like other fundamental abstractions for energy-efficient computing, search trees need to support ... more Like other fundamental abstractions for energy-efficient computing, search trees need to support both high concurrency and finegrained data locality. However, existing locality-aware search trees such as ones based on the van Emde Boas layout (vEB-based trees), poorly support concurrent (update) operations while existing highly-concurrent search trees such as the non-blocking binary search trees do not consider data locality. We present GreenBST, a practical energy-efficient concurrent search tree that supports fine-grained data locality as vEB-based trees do, but unlike vEB-based trees, GreenBST supports high concurrency. GreenBST is a k-ary leaf-oriented tree of GNodes where each GNode is a fixed size tree-container with the van Emde Boas layout. As a result, GreenBST minimizes data transfer between memory levels while supporting highly concurrent (update) operations. Our experimental evaluation using the recent implementation of non-blocking binary search trees, highly concurrent B-trees, conventional vEB trees, as well as the portably scalable concurrent trees shows that GreenBST is efficient: its energy efficiency (in operations/Joule) and throughput (in operations/second) are up to 65% and 69% higher, respectively, than the other trees on a high performance computing (HPC) platform (Intel Xeon), an embedded platform (ARM), and an accelerator platform (Intel Xeon Phi). The results also provide insights into how to develop energy-efficient data structures in general. 100% 95% 90% 80% 50%
As other fundamental programming abstractions in energy-efficient computing, search trees are exp... more As other fundamental programming abstractions in energy-efficient computing, search trees are expected to support both high parallelism and data locality. However, existing highly-concurrent search trees such as red-black trees and AVL trees, do not consider data locality while existing locality-aware search trees such as those based on the van Emde Boas layout (vEB-based trees), poorly support concurrent (update) operations. This paper presents DeltaTree, a practical locality-aware concurrent search tree that combines both locality-optimisation techniques from vEB-based trees and concurrency-optimisation techniques from non-blocking highly-concurrent search trees. DeltaTree is a k-ary leaf-oriented tree of DeltaNodes in which each DeltaNode is a size-fixed tree-container with the van Emde Boas layout. The expected memory transfer costs of DeltaTree's Search, Insert and Delete operations are O(log B N), where N, B are the tree size and the unknown memory block size in the ideal cache model, respectively. DeltaTree's Search operation is wait-free, providing prioritised lanes for Search operations, the dominant operation in search trees. Its Insert and Delete operations are non-blocking to other Search, Insert and Delete operations, but they may be occasionally blocked by maintenance operations that are sometimes triggered to keep DeltaTree in good shape. Our experimental evaluation using the latest implementation of AVL, red-black, and speculation friendly trees from the Synchrobench benchmark has shown that DeltaTree is up to 5 times faster than all of the three concurrent search trees for searching operations and up to 1.6 times faster for update operations when the update contention is not too high.
Locust beans (Parkiabiglobosa/Filicoidea) and Tamarind (Tamarindusindica) are trees grown in most... more Locust beans (Parkiabiglobosa/Filicoidea) and Tamarind (Tamarindusindica) are trees grown in most tropical region. They are multipurpose trees with fruits that play a vital role most especially in preparing food delicacy, drinks and commercial purpose as well as for medical values. The fruits pods of locust beans and tamarind trees are normally discarded after opening to remove the seeds, powder and the pasty parts of the fruits for use. The pods when soaked in water releases brownish/pinkish colour which is harmful to aquatic life. The synthetic dyes/colours also poses a threats to human life, pollutes our environment The synthetic dyes/colorant poses a threats to human life, pollutes our environment and this coupled with Nigerian problems of scarcity of dyes, cost of importation, environmental pollution and unemployment opportunities. Therefore, the international community called for urgent attention, urging nations to go back to the re- used of natural dyes to served as an alternatives to this menaced. In order to find a lasting solution to environmental pollution caused by the discarded pods and also create alternative to the harmful synthetic dyes; this study researches on the potential of using the brownish/pinkish colour released by the pods as textile dyes.
This study assesses the effectiveness of managers of technical college workshops from the perspec... more This study assesses the effectiveness of managers of technical college workshops from the perspective of stakeholders in North Central Nigeria. A descriptive survey design was employed. Three research questions and two hypotheses were formulated to guide the study. A 35-item questionnaire developed by the researchers and validated by three experts was used to collect data from 101 administrators, 140 teachers, and 24 workshop personnel randomly sampled and stratified along trades in technical colleges. Mean was employed to answer the research questions while one way analysis of variance (ANOVA) was employed to test the hypotheses using Statistical Package for Social Sciences (SPSS) for data analysis. Result revealed that respondent shared similar opinions on the management of technical college workshops. It was therefore recommended that: The managers of the workshops should use alternative approaches in the management of workshop instead of waiting for government to provide everyth...
Proceedings of the 2015 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems, 2015
Like other fundamental abstractions for high-performance computing, search trees need to support ... more Like other fundamental abstractions for high-performance computing, search trees need to support both high concurrency and data locality. However, existing locality-aware search trees based on the van Emde Boas layout (vEB-based trees), poorly support concurrent (update) operations. We present DeltaTree, a practical locality-aware concurrent search tree that integrates both locality-optimization techniques from vEB-based trees, and concurrency-optimization techniques from highly-concurrent search trees. As a result, DeltaTree minimizes data transfer from memory to CPU and supports high concurrency. Our experimental evaluation shows that DeltaTree is up to 50% faster than highly concurrent B-trees on a commodity Intel high performance computing (HPC) platform and up to 65% faster on a commodity ARM embedded platform.
2009 6th International Symposium on High Capacity Optical Networks and Enabling Technologies (HONET), 2009
Abstract The aim of this work is how to speed up the process of the biological (DNA and proteins... more Abstract The aim of this work is how to speed up the process of the biological (DNA and proteins) sequence comparison process by using a hybrid parallelisation technique of combining different parallel methods. Smith-Waterman algorithm has been known as the most ...
2010 Second International Conference on Computer Research and Development, 2010
AbstractCurrent work on bipartite graph-based algorithm for protein tertiary structure matching ... more AbstractCurrent work on bipartite graph-based algorithm for protein tertiary structure matching shows that the algorithm demands heavy computation and extensive processing time during graph preparation and matching. In this work, we deployed multithreading approach ...
2008 Second Asia International Conference on Modelling & Simulation (AMS), 2008
We proposed a parallel linear storage space with a sub linear complexity text searching algorithm... more We proposed a parallel linear storage space with a sub linear complexity text searching algorithm by a combination of an offline and online text searching methods. We achieved this by modeling a Semi-self Index which is consists of text along its index in an efficient and compact ...
Environmental Management and Sustainable Development, 2013
This study investigates the extraction of colour from locust beans fruits pods to dye cotton and ... more This study investigates the extraction of colour from locust beans fruits pods to dye cotton and silk fabrics using the mordanting treatment with these mordants of CUSO4, FERROUS (II) & (III) SULPHATE AND ALUM. Aqueous and solvent extraction method was adopted for this experiment. The colour obtained was different shades ranging from soft, light brown and brown, depending on the mordant used. However, the change in colour can be improved using other mordant from the ones used in this study, fastness properties test was carried out and proved good result. Therefore, the findings will benefit local dyers, farmers, textiles industries; organisations, institutions that deal with dyes/colour, research institutes and can be good for future research on other natural fabrics.
Journal of Radioanalytical and Nuclear Chemistry, 1999
The leaching of137Cs and90Sr from Surrey Earth montmorillonite and Silver Hill illite has been st... more The leaching of137Cs and90Sr from Surrey Earth montmorillonite and Silver Hill illite has been studied. Three forms of the clays, calcined and uncalcined and their composites with cement, were leached with four leachants, deionized water, and simulated sea ground and storage pond waters. Results were expressed as the cumulative fraction of activity leached against the square root of time, and used to compute diffusion coefficients. These suggested that all the leaching processes studied were slow (10−22 to 10−18 m2.s−1).
Journal of Materials Chemistry, 2000
J. Mater. Chem., 2000, 10, 2734-2740 DOI:10.1039/B006662L (Paper). The uptake of caesium and stro... more J. Mater. Chem., 2000, 10, 2734-2740 DOI:10.1039/B006662L (Paper). The uptake of caesium and strontium radioisotopes onto clays. Allen Dyer*, John KK Chow and Ibrahim M. Umar. Science Research Institute, Chemistry ...
2022 IEEE Delhi Section Conference (DELCON)
Recently, human activity recognition (HAR) and classification have driven interest in the industr... more Recently, human activity recognition (HAR) and classification have driven interest in the industry and academic studies due to a huge application in the area. However, monitoring of activity from video and images is challenging due to the nature of video data, different weather conditions, the uncertain movement of human being, and availability of suitable datasets. There is very few research on deep learning-based activity recognition as the area is in its early stages. However, existing study was able to detect single individual known activities but unable to detect unknown and complex activities. This can be extended from single person to multi-person classification. Additionally, in the existing 1Dimentional Convolutional Neural Network (1D CNN), the core goes in a single way, thus, the input and output data of 1D CNN is 2-dimensional. Therefore, this technique is more suitable for time series data than images or videos. However, optical flow is a best way to abstract and define shadows for an active video background. Hence, this research further explores activity recognition using multiple activities frames, optical flow, I3D CNN, LSTM and SoftMax Classifier. In this work, LSTM is employed to aid in detecting information based on HAR recognition. The i3D-CNN model, optical flow approach and LSTM algorithm are employed in the feature extraction step. Experimental results on MATLAB 2021a shows that, improving the number of hidden layers more than the number of layers can decrease the accuracy in the test set, thereby making the network to fit to the training set. It was further observed that the proposed model achieved best and stable recognition accuracy when the configuration settings were 10 LSTM layers and 256 hidden units. In this case the proposed model achieved the best-case scenario for the recognition accuracy with 97.8% for draw sword, 94.8% for shoot bow, 9.81% for shoot gun, 94.85% for Hit and 95.9% for punch respectively. On comparison with other base line classifiers, the SoftMax classifier attained the best recognition accuracy of 97.8% as against the SVM with 94.35% and KNN with 92.91%.
Uploads
Papers by Ibrahim Muhammad Umar