Immediate availability of a new Convolutional Neural Network has been announced by Omnitek, delivering high performance per watt at full FP32 accuracy in a midrange SoC FPGA. Optimised for the Intel Arria 10 GX architecture, the Omnitek Deep Learning Processing Unit (DPU) achieves 135 GOPS/W at full 32-bit floating point accuracy when running the VGG-16 CNN in an Arria 10 GX 1150.

Specialist in FPGA IP, Omnitek, has announced immediate availability of a unique 3D LUT for colour space conversions and conversion between nonlinear gamuts. This is a critical function in high-demand applications such as Rec. 709/2020 and SDR/HDR conversions, chroma keying and artistic effects such as sepia, black-and-white or vivid colour.

Omnitek is spinning out its inhouse expertise in high speed video interfacing technologies to enable manufacturers to speed up their own designs and catch the wave of UHD without breaking the bank on expensive development teams and long timescales. With off the shelf IP-cores and custom designs for blue-chip clients, Omnitek has years of experience developing firmware and design-for-manufacture PCBs.

As high resolution cameras and displays spread to medical, AV, events, military and telecomms applications, the need for practical, real world video processing solutions has grown. Simultaneously, FPGA processing capacity has grown to allow more ambitious designs. Omnitek provides off-the-shelf and bespoke designs for video equipment manufacturers, allowing them to focus on USPs without needing expensive R&D teams.