About
Contact us
Career

Home
Software development
L10N and I18N
Testing and QA
R&D
Products
Company

About
Contact Us
Career

Localization and
Internationalization

Localization Engineering
Translation & Linguistic Review
Desktop Publishing

Software Testing
and QA

Manual Testing
Automated Testing
Mobile Software Testing

CUDA and High
Performance Computing

High Performance Computing
CUDA

Research and
Development

Hadoop
CUDA

Discover product

Software Statistics
IT Information System
Discover product

Doctor ELEKS
Medical Information System
Discover product

Dakar
Power System Analysis
Discover product

Screen Viewer
Quality Assurance Tool
Discover product

eViewer
Nice Preview Tool
Discover product

eLinkChecker
Perfect Localization Tool
Discover product

My Fonts
Font Management System

ACCELERATED IMAGE PROCESSING

NVIDIA® CUDA™ SOLUTION FOR HEALTHCARE

Read more about this SVD acceleration project where the maximum achieved performance improvement is 155x!

View case study

Natural Disaster Impact Forecasting Model

NVIDIA CUDA Solution

View case study

CUDA-Based Financial Risk Simulation

Financial Risk Simulation

Speaking in London

Strategic Testing Approach Presentation by Taras Tovstyak

C++ developer

(Graphics programming)

The Story About ELEKS.COM

Corporate Website

Looking for the newest technological trends and proven approaches with the creativity and bright ideas? New, bright, responsive and in the cloud! The new ELEKS website!

View case study

<- Back

Accelerated Image Processing

NVIDIA® CUDA™ Solution for Healthcare

ELEKS helps a medical device company reduce patient assessment time by accelerating Magnetic Resonance Imaging (MRI) scanner post-processing software with CUDA™

Customer

US MRI scanners developer and manufacturer

Industry

Healthcare

Challenge

Reduce patient time on the table by accelerating complex MRI image post-processing and reconstruction tasks.

Solution

ELEKS developed a library of specialized functions to perform extremely fast and efficient MRI images reconstruction and SVD processing using NVIDIA® CUDA™. Now these operations take less than a second compared to 28 seconds for FFT and 18.5 seconds for SVD processing (over benchmark datasets using single-threaded reference CPU implementation). And the maximum achieved SVD acceleration is 155x on a benchmark dataset of 96 512x384 frames acquired from 8 coils in YZ orientation.

Business Benefits

Reduced patient time on the table
Reduced patient assessment duration
Increased clinics revenues from using MRI scanners
Reduced development time and cost

Technologies and tools

NVIDIA® CUDA™ GPUs, C++, Microsoft .NET, C#.

Solution Overview

Parallel 2D & 3D Inverse FFT on CUDA™.
Image reconstruction using 2D & 3D inverse Fast Fourier Transform is implemented using NVIDIA® cuFFT library. Special consideration was given for the large data frame cases, when there is not enough memory on-board to perform FFT. To overcome this limitation ELEKS engineers implemented a mathematically correct algorithm to run FFT over portions of the large data frame with subsequent aggregation of partial results.
High-performance batch SVD processing on CUDA™.
SVD is a largely iterative algorithm, and parallelization of a single iteration does not give any significant benefits, especially for smaller image sizes. Therefore, the only realistic way to accelerate this operation is to run SVD processing over all frames and coils in parallel. Unfortunately, at that time no libraries on the market provided suitable SVD implementation with batching. ELEKS developers solved this problem by porting SVD algorithm to CUDA and running it over all frames and coils in parallel.

Download PDF

Advantages

Less than a second processing time for MRI image post-processing operations
High-quality software fully reconciled and ready to use
Hardware selection recommendations backed by experience and extensive benchmarks
Complete API reference

Scope of Services

Analysis of the problem and available solutions on the market
Mathematical research and development of algorithms
Review and analysis of existing source code and libraries
Implementation of CUDA™ code, C++ library and .Net wrapper
Visual and binary reconciliation of final images
Benchmarking of CUDA™ implementation on various NVIDIA® CUDA™ GPUs.
Consulting as to the applicability of CUDA™ and selection of hardware.

<- Back

NATURAL DISASTER IMPACT FORECASTING MODEL

IMPROVED THE APPLICATION PERFORMANCE BY OVER 100 TIMES

ELEKS developed a solution for forecasting and evaluation of losses after natural disasters using Monte-Carlo simulation. Using CUDA™ has significantly improved runtime and allowed the use of more Monte Carlo paths for better convergence.

Customer

A leading provider of risk management services, insurance and reinsurance brokerage

Industry

Insurance

Business Challenge

To provide consulting with regard to the applicability of CUDA™ for impact forecasting models and optimal ways of building GPU-enabled application to reach optimum performance.
To improve the runtime of the application at least 15 times by moving the computation module to GPGPU, particularly, NVIDIA® CUDA™ cards.

Technologies

NVIDIA® CUDA™, Microsoft .Net

Services ELEKS has provided

Performed extensive benchmark of original and CUDA™ code
Restructured the original code to isolate the computation model and reduce memory footprint
Implemented the computation model in CUDA™ C++ API to achieve the maximum possible speed improvement
Provided CPU fallback mode to run the model on CPU if GPU is not available
Investigated the tools available on market to generate CUDA™ code from .NET byte code to see if the requirement to remain within .NET space can be fulfilled

Download PDF

As a result

Using CUDA™ for impact forecasting model improved execution time 28.5 times on 25 Monte- Carlo paths and 166 times on 200 paths
Provided detailed benchmarking and cost benefit analysis comparing unmodified code execution vs. CUDA™ vs. CPU fallback mode on various types of hardware

Business benefits

The performance of the application was improved by over 100 times and exceeded all customer’s expectations
Speeding time for processing significantly reduced the cost

<- Back

NEAR REAL-TIME CUDA-BASED FINANCIAL RISK SIMULATION

FAR OUTPERFORMS TRADITIONAL LARGE SCALE COMPUTING ENVIRONMENTS

Thanks to the solution developed by ELEKS, information about financial risks of insurers is actualized within minutes, compared to days, when previous generation systems were used. The benefit is that the computing horsepower far outperforms traditional large scale computing environments, delivering nearly real-time results instead of those that previously took several hours to produce a single report.

Customer

A leading provider of risk management services, insurance and reinsurance brokerage.

Industry

Banking & Financial, Insurance

Business Challenge

To reduce the companies’ risk caused by market downturns by providing real-time evaluations of a hedge fund portfolio.

Solution

ELEKS developed an NVIDIA® CUDA-based solution as determined by requirements from a Fortune 500 customer. This particular CUDA™ solution runs Monte Carlo simulations in a High Performance Computing environment to calculate market parameters in real-time. The current state of the entire portfolio is monitored at a rate never achieved before and is continuously made available to traders. The models can be fine-tuned by analysts using easy-to-use authoring tools.

Technologies

NVIDIA® CUDA™
Microsoft .NET for backend services
Windows Presentation Foundation (WPF) for user interface
Python with numerical libraries
Red Hat Enterprise Linux
Citrix XenApp and XenServer virtualization

Solution Overview

ELEKS refined the architecture of the system, plus designed and implemented front-end and back-end components.

The CUDA-enabled solution developed by ELEKS is running Monte Carlo simulations in a massively parallel computation environment across an HPC grid with 17,920 CUDA™ cores in total. Should extra processing power be needed, additional CUDA-enabled servers can be added to expand the array on this truly scalable system. This innovative solution drastically reduces the cost for complex computing. As compared to the traditional high-end computing environment, due to higher density and higher parallelization amount, it slashes the cost for power and rack space.

As a result

Risk and market evaluation with Monte-Carlo simulations developed
Ability to evaluate complex models and analytics in a nearly real-time fashion added

Download PDF

Business Benefits

Radical reduction of reaction time to adverse market conditions
Integrated view on hedge performance and risks
Acceleration of decision making process
Short time to market period for new models
Cost reduction

Custom SoftwareDevelopment

Localization andInternationalization

Software Testing and QA

Featured Technologies

CUDA and HighPerformance Computing

Research andDevelopment

Software Statistics

Doctor ELEKS

Dakar

Screen Viewer

eViewer

eLinkChecker

My Fonts

ACCELERATED IMAGE PROCESSING

Natural Disaster Impact Forecasting Model

CUDA-Based Financial Risk Simulation

Speaking in London

C++ developer

The Story About ELEKS.COM

Accelerated Image Processing

NVIDIA® CUDA™ Solution for Healthcare

Customer

Industry

Challenge

Solution

Business Benefits

Technologies and tools

Solution Overview

Advantages

Scope of Services

NATURAL DISASTER IMPACT FORECASTING MODEL

IMPROVED THE APPLICATION PERFORMANCE BY OVER 100 TIMES

Customer

Industry

Business Challenge

Technologies

Services ELEKS has provided

As a result

Business benefits

NEAR REAL-TIME CUDA-BASED FINANCIAL RISK SIMULATION

FAR OUTPERFORMS TRADITIONAL LARGE SCALE COMPUTING ENVIRONMENTS

Customer

Industry

Business Challenge

Solution

Technologies

Solution Overview

As a result

Business Benefits

Custom Software
Development

Localization and
Internationalization

Software Testing
and QA

Featured
Technologies

CUDA and High
Performance Computing

Research and
Development