While LabVIEW features a rich set of built-in signal processing and analysis functions, new IP and data types are needed as engineers and scientists continue to take advantage of LabVIEW and NI I/O hardware for computationally demanding applications—from nuclear fusion research to signal intelligence and sophisticated robotics systems. In addition, scientific systems that require large amounts of data processing in short time intervals are increasingly combining computing elements including CPUs, field-programmable gate arrays (FPGAs), and graphics processing units (GPUs) in order to take advantage of the respective strengths of each of these devices.
Continue reading to learn about several new additions to the LabVIEW platform: the LabVIEW Multicore Analysis and Sparse Matrix Toolkit, LabVIEW GPU Analysis Toolkit, and stereo binocular vision functions in the NI Vision Development Module.
2. Solve Challenging Problems With the LabVIEW Multicore Analysis and Sparse Matrix Toolkit
LabVIEW has been able to execute multiple algorithms in parallel on multicore processors for some time. However, it has not fully taken advantage of these processors to improve the performance of individual algorithms on single data sets. The LabVIEW Multicore Analysis and Sparse Matrix Toolkit adds more optimal core linear algebra and FFT algorithms based on the multithreading capabilities of the Intel Math Kernel Library (MKL) on the Windows platform. Additional linear algebra functions are also available for LabVIEW Real-Time (ETS) systems that leverage the sequential Intel MKL.
Figure 1. The multithreaded FFT built into the LabVIEW Multicore Analysis and Sparse Matrix Toolkit shows a 3-4x improvement over the standard single-threaded FFT when run on a quad-core processor.
Furthermore, the matrices found in various engineering and science applications often include many zero-valued elements. Linear algebra operations on these sparse matrices can exploit this property to make faster computations on much larger matrices. The toolkit adds new classes, VIs, and graphs for working with real and complex sparse matrices on both Windows and LabVIEW Real-Time (ETS) targets. Finally, all of the new libraries in the toolkit operate on both double- and single-precision floating-point data, so you can optimize the balance between data precision, computation performance, and memory usage
3. Communicate Between LabVIEW Applications and GPU Applications With the LabVIEW GPU Analysis Toolkit
Originally designed to support graphics processing operations, GPUs feature a much different hardware architecture than CPUs—with a number of limited-capability floating-point processing cores often in the hundreds. Due to this highly parallel nature, GPU hardware is being used more and more for performing general-purpose computation on highly parallel data sets.
With the LabVIEW GPU Analysis Toolkit, you can communicate between LabVIEW desktop applications and NVIDIA CUDA GPU hardware. In addition, you can execute a set of prewrapped signal processing and linear algebra functions on the GPU (including selected CUBLAS, CUFFT, and other libraries), or wrap your own GPU applications and call them from LabVIEW.
4. Expand Your Vision System Capabilities With New Binocular Stereo Vision Functions in the NI Vision Development Module
Vision is an integral and essential component for a wide range of test and control applications. Many such applications have the potential to greatly benefit from 3D space information, which can simplify and improve vision system accuracy and reliability. For example, applications in which 3D vision information is useful include palletizing and depalletizing; precision robotic component assembly; pick-and-place machines; surface measurements; inspection systems; and surveillance systems.
Figure 2. Binocular stereo vision functions available in the NI Vision Development Module enable you to extract 3D information from a set of images.
Binocular stereo vision systems involve two cameras mounted almost parallel to each other with a short distance between them. With the latest version of the Vision Development Module, you can now extract and benefit from 3D information in your LabVIEW applications by taking advantage of a rich set of binocular stereo vision functions.
5. Learn More About New LabVIEW Capabilities or Download Evaluation Software
Use the following links to learn about other new capabilities in LabVIEW, or download evaluation versions of the software referenced on this page.