Omnitek achieves world-leading CNN performance per watt in a midrange programmable device

BASINGSTOKE, UK. April 16th, 2019 – Omnitek today announced immediate availability of a new Convolutional Neural Network, delivering world-leading 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.  The innovative design employs a novel mathematical framework combining low-precision fixed point maths with floating point maths to achieve this very high compute density with zero loss of accuracy.

Scalable across a wide range of Arria 10 GX and Stratix 10 GX devices, the DPU can be tuned for low cost or high performance in either embedded or data centre applications.  The DPU is fully software programmable in C/C++ or Python using standard frameworks such as TensorFlow, enabling it to be configured for a wide range of standard CNN models including GoogLeNet, ResNet-50 and VGG-16 as well as custom models.  No FPGA design expertise is required to do this.

Roger Fawcett, CEO at Omnitek, commented “We are very excited to apply this unique innovation, resulting from our joint research program with Oxford University, to reducing the cost of a whole slew of AI-enabled applications, particularly in video and imaging where we have a rich library of highly optimised IP to complement the DPU and create complete systems on a chip.”

FPGAs are being adopted as the platform of choice for many intelligent video and vision systems.  They are ideally suited to Machine Learning applications due to their massively parallel DSP architecture, distributed memory and ability to reconfigure the logic and connectivity for different algorithms.  To this latter point, Omnitek’s DPU can be configured to provide optimal compute performance for CNNs, RNNs, MLPs and other neural network topologies which exist today and more importantly, the as yet unknown algorithms and innovative optimisation techniques that will exist in future due to the significant research in this field.

More information is available at www.Omnitek.tv/dpu.

About Omnitek

Omnitek is a world leader in the design of intelligent video and vision systems based on programmable FPGAs and SoCs. Through the supply of expert design services with highly optimised FPGA intellectual property cores covering high-performance video / vision and AI / machine learning, Omnitek can provide cost-optimised solutions to a broad range of markets. To complement this business Omnitek also designs and manufactures a comprehensive suite of video test & measurement equipment.



Contact:

Giles Peckham
Email Contact

Featured Video
Jobs
Principal Engineer for Autodesk at San Francisco, California
Senior Principal Software Engineer for Autodesk at San Francisco, California
Machine Learning Engineer 3D Geometry/ Multi-Modal for Autodesk at San Francisco, California
Mechanical Engineer 3 for Lam Research at Fremont, California
GIS Specialist for Washington State Department of Natural Resources at Olympia, Washington
Senior Principal Mechanical Engineer for General Dynamics Mission Systems at Canonsburg, Pennsylvania
Upcoming Events
Digital Twins 2024 at the Gaylord National Resort & Convention Center in, MD. National Harbor MD - Dec 9 - 11, 2024
Commercial UAV Expo 2025 at RAI Amsterdam Amsterdam Netherlands - Apr 8 - 11, 2025
Commercial UAV Expo 2025 at Amsterdam Netherlands - Apr 8 - 10, 2025
BI2025 - 13th Annual Building Innovation Conference at Ritz-Carlton Tysons Corner McLean VA - May 19 - 21, 2025



© 2024 Internet Business Systems, Inc.
670 Aberdeen Way, Milpitas, CA 95035
+1 (408) 882-6554 — Contact Us, or visit our other sites:
TechJobsCafe - Technical Jobs and Resumes EDACafe - Electronic Design Automation GISCafe - Geographical Information Services  MCADCafe - Mechanical Design and Engineering ShareCG - Share Computer Graphic (CG) Animation, 3D Art and 3D Models
  Privacy PolicyAdvertise