An SDK option for performing computational processing of data acquired by a digitizer at supercomputer-level speeds using Nvidia's CUDA.
Spectrum Instrumentation GmbH in Germany released a new SDK, the SCAPP software option – the Spectrum CUDA Access for Parallel Processing – on November 20, 2017. SCAPP is an SDK option that uses Nvidia's CUDA standard (CUDA is a parallel computing architecture that utilizes GPUs with 256 to 5,000 processing units provided by NVIDIA) to perform computational processing of data acquired by digitizers at supercomputer-like speeds. With CUDA, it is possible to execute data conversion, filtering, accumulation and averaging, baseline correction, window processing, and FFT at supercomputer-like speeds (up to 12 TFLOPS). The SCAPP package is an extension of drivers that enable Remote Direct Memory Access (RDMA), and it supports the M4i series and M2p series under Linux.
Inquire About This Product
basic information
The SCAPP software option includes the following: 1. Driver extension for direct data transfer (RDMA) between the digitizer and GPU 2. Sample program for data transfer between the digitizer and GPU 3. Sample program for data processing on the GPU SCAPP supports the M4i series and M2p series under Linux. The sample programs are based on C/C++.
Price range
P2
Delivery Time
※About 6 weeks
Applications/Examples of results
Data conversion Filtering Integration and averaging processing Baseline correction Window processing FFT
catalog(8)
Download All Catalogs
News about this product(6)
Company information
Electronics Co., Ltd. IMT Division specializes in the individual sales of high-speed digitizers, AWGs (Arbitrary Waveform Generators), and digital I/O from the German company Spectrum, as well as the development and sales of integrated measurement systems and streaming systems using these products. With our extensive knowledge of products and a robust support system, which is possible because we specialize in high-speed digitizers, we propose comprehensive measurement systems that cater to various applications.