MENU

OpenCL platform for high-performance FPGA accelerator

OpenCL platform for high-performance FPGA accelerator

By eeNews Europe



The ready-to-use software modules in OpenCL assist in the building of custom engines for compute-intensive workloads in applications such as image processing and facial recognition. OpenCL is a standardized, open environment for parallel programming of heterogeneous computing systems, that uses standard programming languages and contains APIs to allow the qick completion of projects using familiar techniques. The new AI platform includes a compiler to allow users to build their own applications.

The modules are preloaded on the accelerator card and currently include vector addition, FFT and 2-dimensional FFT, edge detection, file transcoding, face detection, and Sobel edge detection. Arrow’s team will continue to create new functions and is in the process of integrating popular AI frameworks including Caffe and TensorFlow, leveraging OpenCL.

BittWare 385A features the Intel Arria 10 1150 GX FPGA connected to the host via PCIe Gen 3 x8 and directly to two 40Gb Ethernet QSFP+ optical interfaces. Built for high throughput, the card has two banks of 4-16GB DDR3 capable of up to 2133 MTransfer/s. The Arria 10 GX also features heterogeneous acceleration, combining hard floating-point DSP blocks with programmable logic fabric.

All demos loaded on the accelerator card and supported by BittWare’s OpenCL HPC Board Support Package (BSP) and Intel FPGA OpenCL SDK. Camera RealSense is also loaded and ready to use out of the box.

Arrow is providing easy access to the new AI platform through its unique Testdrive program.

More information

www.FiveYearsOut.com

www.arrow.com

Related news

Arrow launches FPGA design competition in Europe

Teltonika IoT Solutions in EMEA | Arrow

Edge computing box offers high virtual machine density

ADLINK and Lenovo team up for IoT and OEM solutions

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News

Share:

10s