Nanoveu (ASX: NVU) has completed comprehensive industry-standard benchmark testing for the ECS-DOT chipset developed by semiconductor business Embedded AI Systems (EMASS) in the categories of anomaly detection and keyword spotting.

The results validate the chipset’s potential to handle computationally intensive, energy-efficient tasks such as 2D-to-3D data conversion based on eye tracking—crucial for real-time adaptive user interactions.

The chipset demonstrated superior energy efficiency, outperforming leading competitors and peers including Qualcomm and Himax, positioning the company as an industry frontrunner in ultra-low-power embedded artificial intelligence (AI) applications.

Anomaly detection

Anomaly detection is an AI and data analysis technique used to identify data points, patterns or behaviours that significantly deviate from what is considered “normal” or expected.

It is valuable in fields where detecting rare or unexpected events is essential for system reliability, security or personalisation, enhancing accuracy and responsiveness.

The ECS-DOT chipset achieved an execution time of 1.22 milliseconds with 0.8 microjoules of energy per inference in anomaly detection testing, representing an improvement of up to 287 times over the best public results reported by global semiconductor company STMicroelectronics.

Keyword spotting

Keyword spotting is a fundamental component for interactive language models that enables a device to “wake up” in response to specific speech triggers, activating more complex applications only when needed.

This approach allows sophisticated systems to remain idle and conserve energy when not in use instead of continuously running background processes.

In keyword spotting testing, the ECS-DOT chipset achieved an execution time of 3.9 milliseconds and 3.07 microjoules of energy per inference, delivering energy efficiency up to ten times that of similar tests conducted on powerful Core chips designed by AI specialist Syntiant.

‘Drive for innovation’

System-on-a-chip (SoC) semiconductor business EMASS founder Mohamed M Sabry Aly expressed pride in the benchmark test performance of the ECS-DOT chipset.

“These results highlight our continuous drive for innovation in ultra-low-power AI solutions,” he said.

“By achieving such energy-efficient performance without compromising accuracy, we are opening new possibilities for embedded AI applications across various industries.”

EMASS acquisition

Nanoveu secured an agreement last month to acquire EMASS in return for $5 million in shares plus 83.3 million performance rights.

The acquisition will enable the integration of EMASS SoC technology with Nanoveu’s proprietary EyeFly3D platform.

The platform combines film and software to transform 2D images on regular digital displays into compelling 3D content without the need for special glasses.