Omnitek Demonstrates Highest Performance Convolutional Neural Network on an FPGA
BASINGSTOKE, UK -- October 2nd, 2018 – Omnitek today announced immediate availability of the highest performance CNN on an FPGA, achieving over 50% higher performance than any competing CNNs and out-performing GPUs for a given power or cost budget. The Omnitek Deep Learning Processing Unit (DPU), available today as a CNN and in future also as an RNN and an MLP, delivers both unrivalled performance on an FPGA and future-proofed performance optimisation, due to the programmable nature of FPGAs. For applications which do not require the ultimate in performance, this performance advantage can be traded for significant reductions in system cost or power.
Demonstrated as a GoogLeNet Inception-v1 CNN, using 8-bit integer resolution, the Omnitek DPU achieves 16.8 TOPS performance and is able to inference at over 5,300 images per second on a Xilinx® Virtex® UltraScale+™ XCVU9P-3 FPGA. This makes it highly suited to object detection and video processing applications at the Edge and in the Cloud, such as intelligent super resolution 8K upscaling for which performance is most important. The DPU is fully software programmable in C/C++ or Python using standard frameworks such as TensorFlow. No FPGA design expertise is required.
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 the wide range of neural network topologies and optimisation techniques which exist today and more importantly, the as yet unknown algorithms that will exist in future due to the significant research in this field.
Roger Fawcett, CEO at Omnitek, commented “We have a roadmap to improve the DPU performance even further through collaborative research with Oxford University on optimisation techniques, alternative network topologies and novel silicon architectures.”
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.
|
Related News
- Omnitek achieves world-leading CNN performance per watt in a midrange programmable device.
- Altera Demonstrates Industry's Highest Performance DDR4 Memory Data Rates in an FPGA
- Altera Demonstrates 90-nm Leadership by Shipping World's Highest-Density, Highest-Performance FPGA
- Synopsys Introduces Industry's Highest Performance Neural Processor IP
- Achronix Now Shipping Industry's Highest Performance Speedster7t FPGA Devices
Breaking News
- Thalia's AMALIA 24.2 introduces pioneering estimated parasitics feature to reduce PEX iterations by at least 30%
- TSMC plans 1.6nm process for 2026
- Qualitas Semiconductor Partners with TUV Rheinland Korea to Enhance ISO 26262 Functional Safety Management System
- M31 has successfully launched MIPI C/D PHY Combo IP on the advanced TSMC 5nm process
- Ceva multi-protocol wireless IP could simplify IoT MCU and SoC development
Most Popular
- Controversial former Arm China CEO founds RISC-V chip startup
- Siemens collaborates with TSMC on design tool certifications for the foundry's newest processes and other enablement milestones
- Credo at TSMC 2024 North America Technology Symposium
- Synopsys Accelerates Next-Level Chip Innovation on TSMC Advanced Processes
- Kalray Joins Arm Total Design, Extending Collaboration with Arm on Accelerated AI Processing
E-mail This Article | Printer-Friendly Page |