# Nanoveu Reports Up to 56 Times Lower Energy per AI Inference in ECS-DoT-22 Benchmarks > Nanoveu's EMASS ECS-DoT-22 chip used up to 56 times less energy per AI inference than published rivals, with evaluation kits now available through DigiKey. - Published: 2026-11-12T01:50:04+00:00 - Section: Research & Clinical Trials - Author: News Desk - URL: https://newsforagent.com/news/2026-11-12/nanoveu-ecs-dot-22-benchmark-energy - Source: [TradingView](https://www.tradingview.com/news/smallcaps:0f494e25a094b:0-nanoveu-reports-up-to-56-fold-lower-energy-use-in-ecs-dot-22-chip-tests/) ## Key takeaways - Nanoveu reported that its ECS-DoT-22 chip used between 17 and 56 times less energy per inference than published results for Syntiant Core 2 and BrainChip's AKD1000. - Keyword spotting on the chip consumed 0.564 microjoules per inference, down from 3.07 microjoules for the 2024 reference model. - ECS-DoT-22 evaluation kits are available through DigiKey, and EMASS expects to release its integrated development environment and AI compiler in November. [Nanoveu](https://akta.pro/company-profiles/00049se/nanoveu) reported that its EMASS subsidiary's ECS-DoT-22 chip used between 17 and 56 times less energy per AI inference than selected published competing devices, according to [TradingView](https://www.tradingview.com/news/smallcaps:0f494e25a094b:0-nanoveu-reports-up-to-56-fold-lower-energy-use-in-ecs-dot-22-chip-tests/). The Australian-listed company said the tests covered keyword spotting, visual wake words, image classification and anomaly detection. ## How the ECS-DoT-22 Benchmarks Were Run Nanoveu tested neural-network models purpose-built for the chip's accelerator. Against its earlier reference models on the same hardware, the company measured processing speed improvements of up to nine times, with higher accuracy across all four workloads. The comparison set included published low-energy results for Syntiant Core 2 on three workloads and BrainChip's AKD1000 on anomaly detection, where the new models used 17 to 56 times less energy per inference. Nanoveu also compared its figures with MLCommons-verified measurements from Asygn's ColibryNPU, finding energy consumption 3.3 to 18 times lower across the four workloads at Asygn's 40MHz operating point. For keyword spotting, the optimised model consumed 0.564 microjoules per inference against 3.07 microjoules for the 2024 reference model, while anomaly detection fell from 0.8 microjoules to 0.385 microjoules. Image classification dropped from 5.5 microjoules to 2.67 microjoules, while visual wake words rose 6% to 3.94 microjoules despite gains in accuracy and speed. At 50MHz, keyword spotting ran nine times faster, image classification 4.5 times faster and anomaly detection three times faster than the reference models. "These models were built for our accelerator, and the difference shows: more accurate than the reference on every workload, up to nine times faster, and keyword spotting and anomaly detection below a microjoule," EMASS founder and Nanoveu director Mohamed Sabry said. ## Where Nanoveu's Edge AI Business Stands Nanoveu, founded in 2012 and headquartered in Singapore, designs ultra-low-power RISC-V edge AI systems-on-chip through [EMASS](https://akta.pro/company-profiles/0006x2a/emass) for always-on inference in wearables, drones, IoT and smart devices, alongside glasses-free 3D screen technology and solar coatings, according to [akta.pro company records](https://akta.pro/company-data). The same records list total funding of $1.9 million, a last round dated May 2025, and a revenue estimate under $1 million. The records also show a January 2026 product launch move covering completion of tape-out of a next-generation 16nm ECS-DoT SoC, approaching fabrication at TSMC with integrated Bluetooth Low Energy, expanded SRAM and floating-point support. ## Nanoveu's Path to the ECS-DoT-22 Tests In May 2026, Nanoveu acquired Singapore's Spinoff Robotics, a drone start-up from SUTD, [Sutd](https://akta.pro/news-signals) reported. In July 2026, Nanoveu requested a trading halt pending results from live drone testing of its ECS-DoT chip, after earlier simulation tests showed a 33% increase in drone flight time, according to AInvest. In August 2026, Nanoveu said it had secured a $300,000 purchase order for its 3D mobile film product, MarketScreener reported. ## DigiKey Kits and the November Software Release EMASS announced DigiKey as global distributor for the ECS-DoT-22 evaluation board on November 10, 2026, EIN Presswire reported. Nanoveu said evaluation kits are available through DigiKey following that distribution agreement, and EMASS expects to release its integrated development environment and AI compiler during November alongside the benchmark models and scripts needed for independent replication. ## Companies in this story - [Nanoveu company profile](https://akta.pro/company-profiles/00049se/nanoveu) · website: http://www.nanoveu.com - [EMASS company profile](https://akta.pro/company-profiles/0006x2a/emass) · website: https://www.emass.ai --- News for Agent · https://newsforagent.com/llms.txt