Automatic Compilation of MATLAB Programs for Synergistic Execution on Heterogeneous Processors

MATLAB is an array language, initially popular for rapid prototyping, but is now being increasingly used to develop production code for numerical and scientific applications. Typical MATLAB programs have abundant data parallelism. These programs also have control flow dominated scalar regions that have an impact on the program's execution time. Today's computer systems have tremendous computing power in the form of traditional CPU cores and throughput oriented accelerators such as graphics processing units(GPUs). Thus, an approach that maps the control flow dominated regions to the CPU and the data parallel regions to the GPU can significantly improve program performance. In this paper, we present the design and implementation of MEGHA, a compiler that automatically compiles MATLAB programs to enable synergistic execution on heterogeneous processors. Our solution is fully automated and does not require programmer input for identifying data parallel regions. We propose a set of compiler optimizations tailored for MATLAB. Our compiler identifies data parallel regions of the program and composes them into kernels. The problem of combining statements into kernels is formulated as a constrained graph clustering problem. Heuristics are presented to map identified kernels to either the CPU or GPU so that kernel execution on the CPU and the GPU happens synergistically and the amount of data transfer needed is minimized. In order to ensure required data movement for dependencies across basic blocks, we propose a data flow analysis and edge splitting strategy. Thus our compiler automatically handles composition of kernels, mapping of kernels to CPU and GPU, scheduling and insertion of required data transfer. The proposed compiler was implemented and experimental evaluation using a set of MATLAB benchmarks shows that our approach achieves a geometric mean speedup of 19.8X for data parallel benchmarks over native execution of MATLAB.

WEPA: Wind energy potential assessment-spatial decision support system

Spatial Decision Support System (SDSS) assist in strategic decision-making activities considering spatial and temporal variables, which help in Regional planning. WEPA is a SDSS designed for assessment of wind potential spatially. A wind energy system transforms the kinetic energy of the wind into mechanical or electrical energy that can be harnessed for practical use. Wind energy can diversify the economies of rural communities, adding to the tax base and providing new types of income. Wind turbines can add a new source of property value in rural areas that have a hard time attracting new industry. Wind speed is extremely important parameter for assessing the amount of energy a wind turbine can convert to electricity: The energy content of the wind varies with the cube (the third power) of the average wind speed. Estimation of the wind power potential for a site is the most important requirement for selecting a site for the installation of a wind electric generator and evaluating projects in economic terms. It is based on data of the wind frequency distribution at the site, which are collected from a meteorological mast consisting of wind anemometer and a wind vane and spatial parameters (like area available for setting up wind farm, landscape, etc.). The wind resource is governed by the climatology of the region concerned and has large variability with reference to space (spatial expanse) and time (season) at any fixed location. Hence the need to conduct wind resource surveys and spatial analysis constitute vital components in programs for exploiting wind energy. SDSS for assessing wind potential of a region / location is designed with user friendly GUI’s (Graphic User Interface) using VB as front end with MS Access database (backend). Validation and pilot testing of WEPA SDSS has been done with the data collected for 45 locations in Karnataka based on primary data at selected locations and data collected from the meteorological observatories of the India Meteorological Department (IMD). Wind energy and its characteristics have been analysed for these locations to generate user-friendly reports and spatial maps. Energy Pattern Factor (EPF) and Power Densities are computed for sites with hourly wind data. With the knowledge of EPF and mean wind speed, mean power density is computed for the locations with only monthly data. Wind energy conversion systems would be most effective in these locations during May to August. The analyses show that coastal and dry arid zones in Karnataka have good wind potential, which if exploited would help local industries, coconut and areca plantations, and agriculture. Pre-monsoon availability of wind energy would help in irrigating these orchards, making wind energy a desirable alternative.

Quality Assurance and Assessment in Technical Education System: A Web Based Approach

Engineering education quality embraces the activities through which a technical institution satisfies itself that the quality of education it provides and standards it has set are appropriate and are being maintained. There is a need to develop a standardised approach to most aspects of quality assurance for engineering programmes which is sufficiently well defined to be accepted for all assessments.We have designed a Technical Educational Quality Assurance and Assessment (TEQ-AA) System, which makes use of the information on the web and analyzes the standards of the institution. With the standards as anchors for definition, the institution is clearer about its present in order to plan better for its future and enhancing the level of educational quality.The system has been tested and implemented on the technical educational Institutions in the Karnataka State which usually host their web pages for commercially advertising their technical education programs and their Institution objectives, policies, etc., for commercialization and for better reach-out to the students and faculty. This helps in assisting the students in selecting an institution for study and to assist in employment.

Enhancing Speedup in Network Processing Applications by Exploiting Instruction Reuse with Flow Aggregation

Instruction reuse is a microarchitectural technique that improves the execution time of a program by removing redundant computations at run-time. Although this is the job of an optimizing compiler, they do not succeed many a time due to limited knowledge of run-time data. In this paper we examine instruction reuse of integer ALU and load instructions in network processing applications. Specifically, this paper attempts to answer the following questions: (1) How much of instruction reuse is inherent in network processing applications?, (2) Can reuse be improved by reducing interference in the reuse buffer?, (3) What characteristics of network applications can be exploited to improve reuse?, and (4) What is the effect of reuse on resource contention and memory accesses? We propose an aggregation scheme that combines the high-level concept of network traffic i.e. "flows" with a low level microarchitectural feature of programs i.e. repetition of instructions and data along with an architecture that exploits temporal locality in incoming packet data to improve reuse. We find that for the benchmarks considered, 1% to 50% of instructions are reused while the speedup achieved varies between 1% and 24%. As a side effect, instruction reuse reduces memory traffic and can therefore be considered as a scheme for low power.

Low cost calibrated mechanical noisemaker for hearing screening of neonates in resource constrained settings

Background & objectives: There is a need to develop an affordable and reliable tool for hearing screening of neonates in resource constrained, medically underserved areas of developing nations. This study valuates a strategy of health worker based screening of neonates using a low cost mechanical calibrated noisemaker followed up with parental monitoring of age appropriate auditory milestones for detecting severe-profound hearing impairment in infants by 6 months of age. Methods: A trained health worker under the supervision of a qualified audiologist screened 425 neonates of whom 20 had confirmed severe-profound hearing impairment. Mechanical calibrated noisemakers of 50, 60, 70 and 80 dB (A) were used to elicit the behavioural responses. The parents of screened neonates were instructed to monitor the normal language and auditory milestones till 6 months of age. This strategy was validated against the reference standard consisting of a battery of tests - namely, auditory brain stem response (ABR), otoacoustic emissions (OAE) and behavioural assessment at 2 years of age. Bayesian prevalence weighted measures of screening were calculated. Results: The sensitivity and specificity was high with least false positive referrals for. 70 and 80 dB (A) noisemakers. All the noisemakers had 100 per cent negative predictive value. 70 and 80 dB (A) noisemakers had high positive likelihood ratios of 19 and 34, respectively. The probability differences for pre- and post- test positive was 43 and 58 for 70 and 80 dB (A) noisemakers, respectively. Interpretation & conclusions: In a controlled setting, health workers with primary education can be trained to use a mechanical calibrated noisemaker made of locally available material to reliably screen for severe-profound hearing loss in neonates. The monitoring of auditory responses could be done by informed parents. Multi-centre field trials of this strategy need to be carried out to examine the feasibility of community health care workers using it in resource constrained settings of developing nations to implement an effective national neonatal hearing screening programme.

Resource Usage Monitoring in Clouds

Monitoring of infrastructural resources in clouds plays a crucial role in providing application guarantees like performance, availability, and security. Monitoring is crucial from two perspectives - the cloud-user and the service provider. The cloud user’s interest is in doing an analysis to arrive at appropriate Service-level agreement (SLA) demands and the cloud provider’s interest is to assess if the demand can be met. To support this, a monitoring framework is necessary particularly since cloud hosts are subject to varying load conditions. To illustrate the importance of such a framework, we choose the example of performance being the Quality of Service (QoS) requirement and show how inappropriate provisioning of resources may lead to unexpected performance bottlenecks. We evaluate existing monitoring frameworks to bring out the motivation for building much more powerful monitoring frameworks. We then propose a distributed monitoring framework, which enables fine grained monitoring for applications and demonstrate with a prototype system implementation for typical use cases.

Automatic compilation of MATLAB programs for synergistic execution on heterogeneous processors

MATLAB is an array language, initially popular for rapid prototyping, but is now being increasingly used to develop production code for numerical and scientific applications. Typical MATLAB programs have abundant data parallelism. These programs also have control flow dominated scalar regions that have an impact on the program's execution time. Today's computer systems have tremendous computing power in the form of traditional CPU cores and throughput oriented accelerators such as graphics processing units(GPUs). Thus, an approach that maps the control flow dominated regions to the CPU and the data parallel regions to the GPU can significantly improve program performance. In this paper, we present the design and implementation of MEGHA, a compiler that automatically compiles MATLAB programs to enable synergistic execution on heterogeneous processors. Our solution is fully automated and does not require programmer input for identifying data parallel regions. We propose a set of compiler optimizations tailored for MATLAB. Our compiler identifies data parallel regions of the program and composes them into kernels. The problem of combining statements into kernels is formulated as a constrained graph clustering problem. Heuristics are presented to map identified kernels to either the CPU or GPU so that kernel execution on the CPU and the GPU happens synergistically and the amount of data transfer needed is minimized. In order to ensure required data movement for dependencies across basic blocks, we propose a data flow analysis and edge splitting strategy. Thus our compiler automatically handles composition of kernels, mapping of kernels to CPU and GPU, scheduling and insertion of required data transfer. The proposed compiler was implemented and experimental evaluation using a set of MATLAB benchmarks shows that our approach achieves a geometric mean speedup of 19.8X for data parallel benchmarks over native execution of MATLAB.

Improving GPGPU Concurrency with Elastic Kernels

Each new generation of GPUs vastly increases the resources available to GPGPU programs. GPU programming models (like CUDA) were designed to scale to use these resources. However, we find that CUDA programs actually do not scale to utilize all available resources, with over 30% of resources going unused on average for programs of the Parboil2 suite that we used in our work. Current GPUs therefore allow concurrent execution of kernels to improve utilization. In this work, we study concurrent execution of GPU kernels using multiprogram workloads on current NVIDIA Fermi GPUs. On two-program workloads from the Parboil2 benchmark suite we find concurrent execution is often no better than serialized execution. We identify that the lack of control over resource allocation to kernels is a major serialization bottleneck. We propose transformations that convert CUDA kernels into elastic kernels which permit fine-grained control over their resource usage. We then propose several elastic-kernel aware concurrency policies that offer significantly better performance and concurrency compared to the current CUDA policy. We evaluate our proposals on real hardware using multiprogrammed workloads constructed from benchmarks in the Parboil 2 suite. On average, our proposals increase system throughput (STP) by 1.21x and improve the average normalized turnaround time (ANTT) by 3.73x for two-program workloads when compared to the current CUDA concurrency implementation.

TCP: thread contention predictor for parallel programs

With proliferation of chip multicores (CMPs) on desktops and embedded platforms, multi-threaded programs have become ubiquitous. Existence of multiple threads may cause resource contention, such as, in on-chip shared cache and interconnects, depending upon how they access resources. Hence, we propose a tool - Thread Contention Predictor (TCP) to help quantify the number of threads sharing data and their sharing pattern. We demonstrate its use to predict a more profitable shared, last level on-chip cache (LLC) access policy on CMPs. Our cache configuration predictor is 2.2 times faster compared to the cycle-accurate simulations. We also demonstrate its use for identifying hot data structures in a program which may cause performance degradation due to false data sharing. We fix layout of such data structures and show up-to 10% and 18% improvement in execution time and energy-delay product (EDP), respectively.

Resource usage monitoring for KVM based virtual machines

Realization of cloud computing has been possible due to availability of virtualization technologies on commodity platforms. Measuring resource usage on the virtualized servers is difficult because of the fact that the performance counters used for resource accounting are not virtualized. Hence, many of the prevalent virtualization technologies like Xen, VMware, KVM etc., use host specific CPU usage monitoring, which is coarse grained. In this paper, we present a performance monitoring tool for KVM based virtualized machines, which measures the CPU overhead incurred by the hypervisor on behalf of the virtual machine along-with the CPU usage of virtual machine itself. This fine-grained resource usage information, provided by the above tool, can be used for diverse situations like resource provisioning to support performance associated QoS requirements, identification of bottlenecks during VM placements, resource profiling of applications in cloud environments, etc. We demonstrate a use case of this tool by measuring the performance of web-servers hosted on a KVM based virtualized server.

Haemophilus influenzae Genome Database (HIGDB): A single point web resource for Haemophilus influenzae

Background: Haemophilus influenzae (H. Influenzae) is the causative agent of pneumonia, bacteraemia and meningitis. The organism is responsible for large number of deaths in both developed and developing countries. Even-though the first bacterial genome to be sequenced was that of H. Influenzae, there is no exclusive database dedicated for H. Influenzae. This prompted us to develop the Haemophilus influenzae Genome Database (HIGDB). Methods: All data of HIGDB are stored and managed in MySQL database. The HIGDB is hosted on Solaris server and developed using PERL modules. Ajax and JavaScript are used for the interface development. Results: The HIGDB contains detailed information on 42,741 proteins, 18,077 genes including 10 whole genome sequences and also 284 three dimensional structures of proteins of H. influenzae. In addition, the database provides ``Motif search'' and ``GBrowse''. The HIGDB is freely accessible through the URL:http://bioserverl.physicslisc.ernetin/HIGDB/. Discussion: The HIGDB will be a single point access for bacteriological, clinical, genomic and proteomic information of H. influenzae. The database can also be used to identify DNA motifs within H. influenzae genomes and to compare gene or protein sequences of a particular strain with other strains of H. influenzae. (C) 2014 Elsevier Ltd. All rights reserved.