Home Update Intel Acquires Omnitek: FPGA Video Acceleration and

Intel Acquires Omnitek: FPGA Video Acceleration and

287


One of the traits of Intel is its funding into new IP. This often takes a number of types, resembling inside R&D, investing in different firms via Intel Capital, or outright acquisitions. Today Intel is saying that’s has acquired Omnitek, a UK primarily based firm that has a portfolio of video acceleration and inferencing IP constructed for FPGAs, in addition to a buyer base constructed on vertical integrations and customized options.

Omnitek, primarily based in Basingstoke UK, has been acquired by Intel. This acquisition consists of all 40 staff of Omnitek, in addition to the HQ, which is able to proceed to be the house of the staff. Omnitek will kind a part of Intel’s Programmable Solutions Group, which covers Intel’s FPGA product line. The worth that Omnitek is about to deliver to Intel’s portfolio revolves round its video acceleration and inferencing IP blocks, at present utilized in projector, audio-visual, Medical, defence, ADAS, safety, and broadcast markets. This consists of over 220 totally different comfortable IP cores that may be applied on Intel’s FPGA pipelines to assist speed up buyer workflow.

Omnitek has a number of angles to help Intel’s increasing portfolio: utilizing FPGAs to enhance video imaging pipelining, resembling pre-processing/filtering of video streams, AI inferencing with DPUs that implement convolutional neural networks optimized for low precision targets, and a customized silicon design and IP options workforce that work with prospects to develop merchandise from the bottom up for particular use circumstances. Intel goals to deliver Omnitek’s choices into its personal for its personal prospects, in addition to make investments right into a roadmap which advantages the portfolio of choices.

Intel at present has numerous IP regarding picture processing and inferencing, such because the Movidius VPU household line, in addition to the push for coaching and inferencing on CPUs and FPGAs. One component to the Omnitek acquisition is to deliver the IP beneath a single umbrella, ideally with a singular software program interface resembling OpenVINO, to permit prospects to have the ability to choose which IP works greatest for his or her situation. By offering a unified software program framework, Intel goals to offer a spread of IP that varies in efficiency, energy, and value, to fulfill any video processing and inferencing want. This has been a typical characteristic of Intel’s latest programmable resolution acquisitions, resembling eASIC, which is already a product choice for Intel’s upcoming Agilex FPGAs.


Omnitek’s Warp FPGA IP, for picture manipulation and overlay as much as 4096×2160 at 60 FPS on 10-bit 4:4:4 

Intel will preserve Omnitek’s merchandise as-is for the close to time period, such because the Flex SoCs which help video warping within the digital signage market, or video stitching applied sciences, or accelerated depth mapping, or Omnitek’s Deep Learning Processing Unit (DPU). As an organization, Omnitek has had a latest progress charge of 30%+ per yr, which Intel believes might be expanded when uncovered to Intel’s buyer base and roadmaps. One component to Omnitek’s worth in line with Intel will likely be within the expanded video area, particularly as video streams contain increased decision content material.

Intel expects to start out providing Omnitek companies and IP to its prospects instantly. The worth of the acquisition was not disclosed.

“Omnitek’s technology is a great complement to our FPGA business. Their deep, system-level FPGA expertise and high performance video and vision related technology have made them a trusted partner for many of our most important customers. Together, we will deliver leading FPGA solutions for video, vision, and AI inferencing applications on Intel FPGAs and speed time-to-market for our existing customers while winning new ones.”

— Dan McNamara, Senior VP, GM, Programmable Solutions Group at Intel

From information facilities to units, compute-intensive purposes like 8K video and synthetic intelligence require a large number of revolutionary compute engines. FPGA units play an more and more essential position, usually…



Source hyperlink

LEAVE A REPLY

Please enter your comment!
Please enter your name here