Run AI applications under 1 mW by skipping zero connections and activations. Deploy as a co-processor chip or tileable IP in an SoC.
Develop your own compressed sparse models in PyTorch and TensorFlow Lite and deploy to the SPU with a Python API.
Sparsified, quantized, fine-tuned, and ready-to-deploy neural networks for applications like speech enhancement, keyword spotting, neural beamforming and more!
Femtosense was at CES 2022 running an ultra-low-latency speech enhancement demo. A video is nice, but hearing is believing.
Reach out to experience our latest version live and in-person. Visit us at CES 2023 if you are around!
Introducing SPU-001, the world's first dual-sparsity AI accelerator for smaller electronic devices. Bring more functionality to your products without affecting battery life or cost.
We’ve built our hardware platform to achieve the efficiency of an ASIC, while retaining the flexibility of a general purpose accelerator. The SPU is easy to program, easy to simulate, and easy to deploy, allowing engineers and product managers to get innovative, class-leading products to market quicker.
10x larger models
100x efficiency
First-class hardware support for dual-sparsity neural networks for low memory footprint and high efficiency.
A wider range of applications and scales
Tileable, all-digital core design with 512 kB per core. Can be ported to different process nodes.
Less time deploying, more time designing
Deploy as a co-processor or IP from PyTorch or TensorFlow. Iterate with the model performance simulator.
Unrestricted to build the impossible
CNNs, RNNs, Transformers, and custom models supported with fine-grain optimization tools.
We’ve built our software development platform to help companies of all sizes deploy optimal sparse AI models for tomorrow’s applications and form factors. Our SDK contains advanced sparse model optimization tools, a custom compiler, and a fast performance simulator. It’s everything you need from exploration to deployment.
Develop and deploy networks from high level frameworks like PyTorch and TensorFlow Lite
Easily prune, quantize, and fine tune sparse models with our model optimization tools
Estimate energy, latency, throughput, and model footprint from a Python API
Notice
This website stores cookies on your computer. These cookies, as specified in our Privacy Policy, are used to collect information about how you interact with our website. We use this information in order to improve and customize your browsing experiences.
Use the “Accept” button to consent to the use of such technologies. Use the “Reject” button to continue without accepting.
We and selected third parties use cookies or similar technologies to store cookies on your computer for technical purposes and, with your consent, for other purposes such as website analytics.
Allow AllThese trackers are used for activities that are strictly necessary to operate or deliver the service you requested from us and, therefore, do not require you to consent.
These trackers help us to measure traffic and analyze your behavior with the goal of improving our service.
Cookies Details