Software architecture specialist Vector Fabrics has launched its Pareon multicore optimization product. Built with an analysis engine capable of preventing what its makers denote as both "hard-to-find" and "hard-to-reproduce" threading bugs, the tool also boasts a hardware modeling engine to prevent the developer from writing code that introduces performance bottlenecks or slowdowns.
Pareon combines the features of Vector Fabrics' previous vfEmbedded and vfThreaded-X86 parallelization tools into one product with over 50 new features, including bus and shared cache contention modeling, support for C/C++, as well as ARM Neon.
NOTE: NEON technology is a 128-bit SIMD (Single Instruction, Multiple Data) architecture extension for the ARM Cortex-A series processors. NEON technology is designed to accelerate multimedia and signal processing algorithms such as video encode/decode, 2D/3D graphics, gaming, audio and speech processing, image processing, telephony, and sound synthesis.
Vector Fabrics' previous tools were cloud-based and ran in the user's web browser, but Pareon runs locally on a workstation, integrating more closely with typical development workflows that include complex build mechanisms, and targets large applications that rely heavily on calling routines inside binary libraries.
"All recent mobile phones, tablets, laptops, and desktops have multicore architectures inside. However, optimizing apps for multicore architectures is hard, error prone, and takes lots of time. As a result, a lot of software isn't optimized yet and dual-core or quad-core devices provide minimal speed-ups. Pareon assists developers in optimizing their software for multicore, allowing them to close this gap and removing the barrier to increase the performance of their applications. We see typical speed-ups of 1.7x on dual-core and 3.5x on quad-core processors," said Mike Beunder, CEO of Vector Fabrics.