22 nm FDSOI · In-Memory Compute · Edge AI

Intelligence
at the EDGE

Advanced AI models face a fundamental memory bottleneck, massive data movement between sensors and processors, that traditional integration cannot handle. NeurIQ solves this with a single transistor that senses, stores, and computes.

Explore Technology → Get in Touch

20 TOPS/W Efficiency

24h Non-Volatile Retention

22nm FDSOI Node

1 Transistor. All Functions

Core Capabilities

💾

Memory

Non-volatile data retention up to 24 hours. Ideal for preloaded AI weights.

⚡

In-Memory Compute

Computation where data lives. Up to 20 TOPS/W with zero data movement.

🔌

FDSOI IP Core

Drop-in compatible with existing low-power FDSOI design flows.

👁️

Image Sensing

Transistor captures light directly.

Core Technology

The Future of Edge AI is
In-Memory Compute

Patented Technology

The Problem

The Memory Bottleneck

Advanced AI models require massive data movement between sensor, memory, and processor demanding excessive power. Traditional module-level integration cannot meet the strict energy-efficiency and latency demands of Edge deployment.

This architectural mismatch between where data is born and where it is processed is the defining constraint of modern Edge AI.

The NeurIQ Solution

Monolithic Silicon Integration

Patented 22 nm FDSOI transistor with added sensing and memory function
Non-volatile storage for up to 24 hours, ideal for preloaded AI weights
Non-floating technology: more attractive than conventional RAM and Flash
FDSOI In-Memory Compute IP core compatible with other low-power devices
Same transistors used for sensing, memory and compute decreasing latency and power usage

Technical Specifications

Feature Performance Capability

Compute Efficiency

upto 20 TOPS/W

Best-in-class Edge AI inference

Power Consumption

order of 100 mW

Full AI Inference Camera module

Data Retention

24 hours

Non-volatile, no refresh required

Process Node

22 nm FDSOI

Monolithic sensor + compute

Operating Modes

Multiple

Image Sensing + Memory + Compute on same device

Integration Type

IP Core

Compatible with low-power FDSOI platforms

Industries & Use Cases

Built for the
Edge of Everything

From autonomous drones to IoT, NeurIQ's technology enables intelligence wherever power and latency budgets are tight.

🚁

Drones & UAV

Onboard AI vision with minimal power usage enabling longer flights and real-time situational awareness.

🤖

Robotics

Low-latency inference for perception and control loops without relying on cloud connectivity.

📷

IoT Cameras

AI-powered edge cameras that process data at the source, reducing bandwidth and ensuring privacy.

📡

RADAR & LiDAR

On-chip compute accelerator for point-cloud processing and signal correlation with FDSOI compatibility.

🛰

GPS Systems

Compute accelerator for positioning algorithms at the silicon level.

⚡

Compute IP Licensing

Embed NeurIQ's In-Memory Compute IP into your own FDSOI chip design for any Edge application.

This is us

The Team

Semiconductor expertise combined with a vision for a fundamentally different approach to Edge AI architecture.

Rob Voorkamp

CEO & Co-founder

Driving business strategy, partnerships, and commercialization of the FDSOI In-Memory Compute IP platform across global markets.

Co-founder & former CEO of SCIL Nanoimprint Solutions
M.Sc. in Mechanical Engineering (TU/e, NL), Executive MBA (TIAS, NL)
>25 years experience in Executive roles at NXP and Philips

LinkedIn rob.voorkamp@neuriq.tech

Deepak Nayak

CTO & Co-founder

Leading architecture and technical development of the In-Memory Compute architecture and IP core design based on patented 22 nm FDSOI transistor.

Experience in semiconductor tool and fab ecosystem
Ph.D. in Nanoscience & Engineering (IISc, IN), Global Executive MBA (INSEAD, FR)
>15 years experience in semiconductor technology

LinkedIn deepak.nayak@neuriq.tech

Intelligenceat the EDGE

The Future of Edge AI isIn-Memory Compute