Reservoir Computer

Making AI’s Expansion Sustainable

Reducing the growing costs of AI models with quantum and photonics.

High-speed affordable AI hardware

Artificial Intelligence (AI) depends on the fast expansion of computing power. Ultimately, data processing speed and computational power limit how intelligent an AI machine can become. 

For example, ChatGPT consumes massive amounts of computing time and network bandwidth to process vast data sets and train its AI models, as well as massive amounts of energy. Its latest version (4) developed by OpenAI, required 100 trillion data parameters, $700K for hosting and serving a day, 34 days and $100M to train, 1287 MWh of electricity and the CO2 emission equivalent of driving 1.4 million miles*.

The first commercial reservoir computer

QCi’s first reservoir computing product is an edge device that is photonic-inspired, FPGA-based, and optimized for recurrent neural network applications.

It’s fast, affordable, energy efficient – and puts the power of a standalone edge computing solution in the palm of your hand, literally. Its advantages over a PC (personal computer) implementation include:

Changing The Equation

QCi is committed to offering sustainable and affordable AI hardware solutions, including the first Reservoir Photonic Computing system, designed to reduce the massive cost, resources, and operation overhead required to train and use neural networks.
Reservoir computing is a best-in-class machine learning algorithm. Our approach enables hardware-based reservoir computing applications on a fast and efficient local device, employing photonic computing and quantum mechanics to deliver the speed and energy efficiencies needed by advanced algorithms and large scale AI models.
These charts represent the accuracy of the QCi device in predicting future events based on past events, using time-series data.

Available Now

The first commercial reservoir computer.

Order Your Reservoir Computer

QCi’s edge device for reservoir computing is shipping later this year. To start the order process, please supply us with your contact information using the form below and we will be in touch with pricing and payment processing information.

Work On My Use Case

QCi partners with application scientists who are applying reservoir computing in real-world use cases. Partners are eligible to receive free computing devices and access, and support to develop targeted use case applications.

Develop Your Reservoir Computing Use Cases Together With QCi

QCi partners with researchers and developers to help advance and develop their reservoir computing use cases. 

Selected partners are eligible to receive free access to QCi reservoir computing devices, as well as user support from QCi application scientist team members. Partners work on their reservoir computing use case application over a targeted time period of 30-90 days, and share findings with QCi. During that time, QCi provides access to internal application scientists through a series of scheduled weekly support sessions. At the end of the period, the Partner and QCi intend to jointly publish non-confidential results from the partnership through a written use case summary or white paper.

If you are already developing use case applications of reservoir computing, reach out to us today to inquire about partnership.


QCi’s Reservoir Computer is ideal for predicting, classifying, and function fitting time-series based data to serve common needs in finance, robotics, weather prediction, sensing, and fraud detection.
Example applications that leverage the power of reservoir computing include:

Speech & Audio Processing

speech recognition, speaker identification, music classification, and confirming the authenticity of human generated video/voice/text

Image & Pattern Recognition

image classification, object recognition, and pattern detection tasks, and ‘noise’ reduction in satellite imagery

Natural Language Processing

language modeling, sentiment analysis, and text classification

Control Systems & Autonomous Vehicles

enhanced decision-making and control processes


QCi Edge Device for Reservoir Computing


200 mm x 100 mm x 55 mm


0.6 kg

Operating Temperature:

room temperature

Power Requirements:

USB-C input power control with 45W power adaptor

Typical Power Consumption:


Data Interface:

10/100/1000 wired ethernet

Transient Response Rate:

10 M nodes/sec

Number of Internal Nodes:

up to 8,192 internal nodes

Operating system:



power on/off

How It Works

QCi’s technology pushes the computing boundaries needed for AI with hybrid architectures that leverage the best of both photonics and electronics data processing.
Combining advanced electro-optics, nanophotonics, analog and digital electronics, QCi’s photonic reservoir system achieves remarkable computing speed and energy efficiency, and has time-frequency multiplexing capability, minimal heat deposition, robustness against RF interference – and the agility to handle Big Data challenges.

Coming Next

The QCi Hybrid Reservoir Computer

This first of its kind device will combine photonic and electronic approaches (opto-electronic) to achieve 100 times greater speeds than digital reservoir computers.


QCi’s plan is to develop AI chips that are available in every data center, to complement GPUs and CPUs, for energy-efficient AI applications.


QCi is already working with partners to deliver business value.

You may also be interested in...

Entropy Quantum Computer

Most Quantum Computers try to exclude noise from the external environment and create perfect systems insulated from nature.
QCi’s Entropy Quantum Computer is uniquely designed to solve for this noise and create useful qubits to perform computations now.

Quantum Photonic Vibrometer

Unlock the future of non-contact material detection with our Quantum Photonic Vibrometer. This revolutionary device, proven to the US Department of Defense, leverages quantum mechanics for unparalleled results, opening new frontiers in remote sensing and beyond.

Contact Us

Fill in the form below to request your reservoir computing devices & access