In CEATEC 2025 in Japan, TDK Corporation presented a prototype that can affect the way artificial intelligence learns and reacts in real time. The new company Analog Reservoir AI Chipdeveloped in collaboration with Hokkaido Universitybrings Biological style and low power learning. to compact hardware. Although it is still a research stage deviceThe prototype vividly demonstrated its potential through an interactive experience: a Rock, paper, scissors game that you can never win..
I tried the demo in person, with a TDK acceleration sensor strapped to my forearm and connected to the prototype chip. While I was preparing to play, the system detected my hand movement almost before movingpredicting my choice with remarkable speed and precision. When I made my gesture, the screen had already shown its winning move.
From digital AI to low-power analog intelligence,
Most AI systems are based on digital computingprocessing large amounts of data through billions of binary operations on dedicated GPUs or accelerators. While powerful, these methods require high energy and cloud resourcespresenting Latency and power restrictions. which make them less practical for compact edge devices such as wearables, sensors or small robots.
TDK’s analog approach is fundamentally different. He Analog Reservoir AI Chip perform calculations through natural dynamics of an analog electronic circuit instead of discrete digital logic. Inspired by him cerebellumthe region of the brain responsible for coordination and adaptation, the circuit can continually learn from feedback — enabling in real time, on the device learning rather than relying solely on pre-trained models.
The underlying concept, known as reservoir computinguses a dynamic system, the “reservoir,” whose internal states evolve in response to input signals. The result is a simple function of those evolving states. Reservoir computing excels in time series processing datasuch as voice, motion or sensor data, because it naturally captures temporal dynamics.
By implementing this framework with analog circuits, TDK eliminates the heavy numerical calculations typical of digital systems. Analog hardware can handle continuous signsreply instantlyand operate with extremely low power consumptionmaking it ideal for real-time learning at the edge.
TDK’s prototype of an analog repository AI chip won an innovation award at CEATEC 2025 – see trophy to the right of the tech spec sheet
Developed with Hokkaido University and inspired by the cerebellum
The prototype was created jointly by TDK and Hokkaido Universitywhose researchers specialize in bioinspired analog computing architectures. The resulting circuit imitates Cerebellar learning and prediction.continually adjusting its internal parameters to align with sensor inputs.
Inspiration comes from cerebellumthe “little brain” located at the base of the human brain. The cerebellum is responsible for Coordination, synchronization and motor learning.continually adjusting movement in response to real-time feedback. Predicts the outcome of an action even before it is completed; for example, adjusting your hand while catching a ball or maintaining balance while walking. TDK’s analog reservoir AI chip reproduces this biological principle in electronic form: continually learns and adaptsusing feedback from the sensor to refine its output almost instantly, just as the cerebellum does with body movements.
Although the prototype is not yet a commercial product, it demonstrates the viability of neuromorphic hardware — Electronics that behave more like biological neurons than traditional processors. TDK foresees potential applications in robots, autonomous vehicles and wearableswhere Adaptability, energy efficiency and instant response. They are crucial.
Recognition at CEATEC 2025
He Analog Reservoir AI Chip received a CEATEC 2025 Innovation Award (Japan Category)recognizing his innovative contribution to real-time edge learning and low power analog computing. The award highlights how TDK’s collaboration with Hokkaido University brings together advanced materials science and neuromorphic circuit design to create practical, energy-efficient AI technology. This distinction underscores the prototype’s potential to transform edge intelligencewhere adaptive learning must occur instantaneously, close to the sensors.
The Rock, Paper, Scissors Demo: AI that Learns You in Real Time
Rock, paper, scissors demonstration at the TDK booth during CEATEC 2025
At CEATEC 2025, TDK presented an engaging demo using its analog tank AI chip and acceleration sensors. The setup included a screen displaying the game, a lightweight sensor on the participant’s arm, and the Chip prototype that processes movement data in real time..When I started moving my fingers to form rock, paper, or scissors, the system measured my finger acceleration and trajectory. He analog circuit instantly processed the data stream and predicted my intended gestureshowing his counterattack before I could finish. The sensation was strange, as if the system had read my mind, but it was responding purely to movement patterns. faster than any human reaction time.
The chip too adapted to my personal movement style. Everyone gestures differently, and when I intentionally changed the way I did the “scissors,” the system I learned the variation on the spot.. Within seconds, he was once again correctly anticipating my movements.
This demonstration highlighted the chip’s strengths:
- Adaptive real-time learning directly from live sensor input
- No cloud connection during operation
- Ultra-low latency and minimal power usage
Hybrid model: cloud calibration and real-time learning at the edge
Although the analog tank AI chip works learning and inference locallyis part of a hybrid AI architecture. According to TDK, large-scale data processing and optimization occur in the cloudwhile individual and real-time learning happens on the edge.
In practice, the chip initial design and calibration were developed using digital simulation toolsprobably in a cloud or in a laboratory environment. The researchers predefined the circuit topology, feedback intensity, and stability parameters. However, once manufactured and working, the chip autonomously adapts to live data without external calculation.
This hybrid model offers the best of both worlds: the cloud provides global optimization and system-level intelligence, while the edge — powered by analog learning — ensures instant response and low power consumption.
Why analog reservoir informatics is important
In AI design, the balance energy efficiency, latencyand learning ability remains a challenge. More current Cutting-edge AI systems executed pre-trained models locally, allowing for fast inference but not continuous learning. Updating these models requires retraining in the cloud, which consumes power and bandwidth.
TDK’s analog reservoir chip changes that paradigm. Because its analog circuits work online learning on devicethey can adapt instantly to new situations: learning from movement, vibration or biosignals without any retraining in the cloud.
This has broad implications for next-generation devices:
- Wearables It could learn a user’s movement or health patterns in real time.
- Robots could autonomously adapt to changing environments.
- Vehicles could continually refine control responses, improving safety and efficiency.
Reservoir informatics aligns perfectly with TDK’s extensive sensor portfoliothat he already drives time series data in motion, pressure, temperature and other domains. Integrating analog AI directly into these sensors could create self-learning components that improve both performance and sustainability.
Motion sensors placed on the thumb and wrist transmitted data to the analog repository’s AI chip, enabling real-time prediction of the user’s hand movement.
The bigger picture: AI in everything, better
TDK’s CEATEC 2025 exhibition focused on the theme of contributing to a “AI ecosystem” – a world where Intelligence is embedded everywhere.from the cloud to the smallest sensor. He Analog Reservoir AI Chip It represents the peripheral layer of this ecosystem, complementing large cloud models rather than replacing them.
Combining cloud-based massive data processing with Individual and adaptive learning at the limit.TDK aims to reduce latency, power consumption and data transmission. This vision aligns with its corporate identity, “In everything, better” reflecting a commitment to incorporating smarter, more efficient intelligence into every product category.
A look at what comes next
While still a prototypehe Analog Reservoir AI Chip shown in CEATEC 2025 provided a clear demonstration of how real-time, low-power learning can take place directly at the edge. Experience has shown that adaptive AI does not require large-scale cloud infrastructure: it can run locally, within an efficient analog circuit.
in it Features sheet displayed at TDK booth (visible in one of our photos), the company listed gesture and voice recognition, anomaly detection and robotics as possible applications. The same sheet highlighted the chip main features: to neural network for modeling time series data, real time learningand low power, low latency operation.
He rock, paper, scissors demonstration It may have been funny, but it showed in a simple way that hardware capable of learning in real time It is no longer a concept: it is already working.
Find more information about TDK Analog Reservoir AI Chip Product Page.
Filed in . Read more about AI (Artificial Intelligence), CEATEC, Chip, Edge, Edge Computing, Japan, Low Power, Processors, Semiconductors and Tdk.
