November 16, 2016
SLX: The next generation of multicore software development tools from Silexica is here today
Automation finally reaches the software development world enhancing productivity and early risk analysis
Saratoga, Calif. – November 16, 2016 – Silexica, the industry leader in multicore software design automation addressing complex, multicore platforms today announced that the next generation of the company’s SLX Tool Suite is shipping now. The SLX Tool Suite has been broadly adopted to automate the distribution of code onto large, multicore platforms. Based upon customer feedback, enhancements were made to this new version, which includes the SLX Parallelizer, SLX Mapper, SLX Generator, and SLX Explorer, resulting in tremendous productivity advancements. The updated tool suite broadens the type and kinds of models, supports additional languages, as well as improves the ease-of-use and early risk analysis.
Building on its capability of exposing parallelism already available in sequential languages like C, the new version of the SLX Parallelizer now adds C++ language support as well. Significant improvements in the ease-of-use of the tool allows the user to simply point to an existing project on disk automating a number of steps that otherwise had to be done manually. This now allows the user to focus their attention on evaluating the tradeoffs and speedup estimations that the tool automatically provides taking into account the number of target cores and on-chip communication of the target platform.
The SLX Mapper and SLX Generator currently solve the software-mapping problem of figuring out and automating which portion of the software should run on which of the available cores. This capability is enhanced in the new version by optimizing this mapping for key system parameters like power and energy in addition to throughput and latency. Target markets like wireless base stations, edge-computing devices and autonomous driving are being constrained by available power considerations and will therefore see benefits by applying these tools in optimizing their systems to meet their power and performance budgets.
"In the near future deep challenges lie in optimizing communication and automotive applications onto multicores. We have successfully integrated the SLX Tool Suite into our multicore software development flow to address these challenges,” said Noriyasu Nakayama, Manager Fujitsu Advanced Technologies Ltd.
As the number of cores in a system continues to rise, the bottlenecks might no longer be in the performance characteristics needed of the cores themselves but rather in the communication between the cores, for example buses, network-on-chip, etc. By modeling this communication, the new release now allows the user to gather key insights into the communication congestion on chip and improve overall system performance of their applications on their multicore platforms. This release also supports 64-bit and PowerPC modeling capabilities for broader platform support.
“We have been doing research based on the SLX Tool Suite for the last four years, first via collaborations with Professor Leupers and his students at RWTH Aachen University, then with the same elite group of experts at Silexica,” said Dr. Xiaotao Chen, Director, Huawei Wireless R&D USA. “The System-level profiling and optimization toolset achieved surprising improvement in both performance and power efficiency over traditional approaches, from processing units (DSP/CPU) power analysis to network-on-chip power efficiency, all the way to power-aware mapping and scheduling of heterogeneous and hierarchical architectures for our wireless baseband applications."
“We believe that as Moore’s Law continues to slow down, and multicores fill the demand gap for computing, the industry must work together to build on common standards and open specifications to accelerate this development. That is why we are contribution our insights and capabilities to open standards like SHIM 2.0 from the Multicore Association, and this current release will cater to these open standards,” CEO Maximilian Odendahl said today.
The SLX Tool Suite is available for immediate download. Please contact firstname.lastname@example.org for an evaluation copy.
About SLX Tool Suite
Today’s software applications addressing complex, advanced products such as wireless baseband processing, edge computing, autonomous driving and embedded vision are driving the demand for more and more processors leading to large multi-core architectures. However, the techniques to efficiently program the heterogeneous multi-core environment largely remains a time consuming, inefficient manual process. Building on multicore research conducted over the last ten years, Silexica began solving this problem with the SLX Tool Suite about two and a half years ago, and we are seeing rapid adoption by hardware and software system architects and designers. Today, the SLX tools SLX Mapper, SLX Generator together deliver fully automated mapping of software tasks and processes onto multicore platforms, optimized for throughput, latency, or power. They leverage native programming toolchains via a source-to-source translation approach to automatically generate architecture aware and middleware-specific C codes. The SLX Explorer guides the selection of the best target hardware by providing insights into software performance/power prediction of the target application. Conversely, the SLX Parallelizer helps customers migrate legacy C, C++ applications into the multicore world giving deep insights into parallelization possibilities based on the target platform.
Silexica is the leading provider of software design automation tools addressing the increasingly complex multicore platforms required to build today’s embedded vision, automotive, edge computing and wireless baseband processing products. SLX tools from Silexica help embedded software developers, system architects and semiconductor providers to differentiate multicore based products. Silexica licenses its technology to renowned electronic companies around the globe and supports its customers from local offices in Germany, U.S., Japan and China. For more on Silexica, visit www.silexica.com