Unlocking the Potential of Today's Multicore Systems
Advancement in modern processor design for the last 10 years has focused on delivering an ever-increasing number of cores. But while most computer systems today include multicore CPUs, most software applications have yet to truly take advantage of the parallel processing potential of these systems. As a result, software applications that used to gain performance with every new processor release have become static, impairing the ability of developers to add new features and functionality, and slowing user adoption of new revisions and new computing systems.
"The lack of multicore programming tools for mainstream developers is the biggest challenge the industry faces today." - Linley Gwennap, Principal Analyst, InfoWorld, March 2009
Texas Multicore Technologies (TMT) unlocks the potential of today's multicore systems by providing revolutionary parallel programming solutions. Based on over 20 years of university research in partnership with NASA, TMT's innovative SequenceL™ technology delivers an automated and agile solution to multicore programming that allows users to specify application functionality without the need to identify parallel structures in the code, fret over race conditions, or be a low level hardware or GPU coding expert. The compiler automatically identifies opportunities for parallel processing in the application and generates multi-threaded code that is transportable across multiple architectures, eliminating the need to rewrite application code for individual multicore computing systems.
Unlike manual brute force parallel coding methodologies, or multicore development environments that merely provide feedback but leave the hard work to the user, TMT provides comprehensive programming solutions that quickly and easily address the root issues facing application developers targeting multicore systems.
- Finding parallelism:
In sequential programming languages, code and data complexity can easily mask opportunities for parallel processing. TMT's solutions automatically find parallelisms from simple, functional descriptions of the application, freeing developers to focus on application functionality and not the parallel implementation.
- System synchronization:
With manual parallel coding, a deep understanding of computing system architecture is necessary to distribute the load across available processors, and to avoid bottlenecks in bus communications and memory accesses. TMT's solutions distribute the application implementation across the available computing resources to optimize application performance.
- Application transportability:
Software applications developed for specific multicore architectures often need to be rewritten to target specific computing systems. TMT's technology enables compilation from implementation-neutral representations of the code. This means that without rewriting the code, the application can be compiled and optimized for other parallel architectures, including: new multicore processors, (GP)GPU cores, and SOCs.
Read on to discover more about how Texas Multicore Technologies can unleash the multicore performance for your existing software applications and accelerate your next development project.
SequenceL™ is a revolutionary parallel programming solution, leveraging the innovative SequenceL functional programming language and powerful auto-parallelizing compiler technology. It greatly simplifies the implementation of multicore applications by eliminating the need to identify and specify parallelism in the code. Just as important, the use of this compact general purpose language and tool chain dramatically cuts code complexity while being essentially self-documenting. This results in faster time to market, improved software quality, and quickly and easily finding opportunities for optimization.
ThreadedS™ is a turn-key software acceleration service using the SequenceL™ solution to upgrade existing applications and new projects for optimized multicore performance.