Enable companies across industries to embed optimized Deep Learning in their innovations using our award winning Deep Learning Optimization Engine and optimally balance the application specific engineering trade-offs between performance, accuracy, energy consumption and cost to help companies create competitive Deep learning based products.
The embedl Deep Learning Optimization Engine
EmbeDL’s award winning technology takes a systematic approach to identify the best methods in the state-of-the-art and to find the optimal way to combine them. In addition, EmbeDL has its own novel methods to optimize surgery on neural networks.
EmbeDL’s technology takes into account the user requirements (such as frames/sec in vision systems) and resource constraints (such as battery power or energy consumption) and the selected hardware platform (CPU/GPU/FPGA and AI accelerators) to optimally compress the neural network as a multi-constrained optimization problem.
EmbeDL’s product creates a complete, fully integrated pipeline from the initial large machine learning model, developed by the customer, to the final compressed model to be deployed in the product. The user simply needs to supply the initial model in any of the major DL frameworks (TensorFlow, PyTorch or Caffe), the requirements and constraints, and EmbeDL will automatically generate the optimized model on a range of possible hardware platforms from which the user can select based on the cost/performance trade-off.
Our award winning technology provides our customers with decreased time to market, decreased project risk, decreased development cost, improved product margins, decreased hardware cost, energy consumption and environmental footprint.
EmbeDL’s technology complements the existing and growing technology ecosystem of both open source projects and proprietary tools from hardware vendors to enable market-ready systems in a time and cost efficient process enabling our customers to develop the next generation of autonomous systems and intelligent applications.
By using state-of-the-art methods for optimizing Deep Neural Networks, we can achieve a significant decrease in execution time and help you reach your real time requirements.
FOOTPRINT IN DEVICE
The EmbeDL Optimization Engine automatically reduces the number of weights , and thus size of the model, to make it suitable to be deployed to resource constraint environments such as embedded systems
The tools are fully automatic, which reduces the need for time consuming experimentation and thus shorter time-to-market. It also frees up your data scientists to focus on their core problems.
Energy is a scarce resource in embedded systems and our optimizer can achieve an order of magnitude reduction in energy consumption for the Deep Learning model execution.
By optimizing the Deep Learning model, cheaper hardware can be sourced that still meets your system requirements leading to improved product margins.
Optimizing and deploying our customers’ Deep Learning models to embedded systems is what we do. By outsourcing this to us, your team can then focus on your core problems.
Interested to learn more?
We would be happy to Talk to you!
We are humbled have been accepted into several prestigious national and international networks bringing together researchers from industry, academia and startups.
We are always flattered when our technology finds itself into media. To reach out with experience from our experimentation, we regularly post findings in our blog. Click here to subscribe.
Join us for this session tomorrow with focus on the innovations of today's industries. You will meet Tanya Marvin-Horowtiz, Partner at Butterfly Ventures, Sandor Albrecht, Senior Project Manager at RISE and EmbeDL CEO and co-founder Hans Salomonsson. More info and...
This is the wonderful coming world of the Artificial-Intelligence-of-Things (AIoT) … with a slight glitch at the end! The Internet of Things has arrived! It is a revolution in technology that will impact us all - both in our professional as well as personal life. The...
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 780681.