Follow Us

ECS-DoT:
Ultra-Low-Power Edge AI SoC
for always on intelligence

ECS-DoT: Ultra-Low-Power Edge AI SoC for always on intelligence

The ECS-Dot is an ultra-low-power, application ready SoC that brings intelligence to the Sensor Edge-where size, power and latency matter most. With dual deep learning accelerators and a novel architecture optimized for highly-compressed models, the ECS-DoT enables real-time processing of sensor, voice and vision data in compact, battery-powered devices-without sacrificing performance or accuracy.

EMASS-I2R
Loading preview

Superior Performance, Lower Power, Small Form Factor

Key Highlights

5mW ultra-low-power
Edge AI processing

Dual DL accelerators
delivering 30 GOPs

32-bit RISC-V core with
floating point support

4MB on-chip memory
(2MB SRAM + 2MB MRAM)

Built-in decompression
IP (~1.3 bits/weight)

Standard I/O: I²C, I²S, UART, QSPI, CPI, 32 GPIOs

5mm x 5mm QFN package

Product Brochure

Competitive Comparison

Compared with leading competitors, ECS-DoT operates up to 93% faster while consuming 90% less energy, all while supporting true multimodal sensor fusion on device

Compared with leading competitors, ECS-DoT operates up to 93% faster while
consuming 90% less energy, all while supporting true multimodal sensor fusion on device

Feature / Metric

EMASS ECS-DoT

Competitor A

Competitor B

Power per Inference

0.1–5 mW

5–100 mW

30–150 mW

Latency

<10 ms

10–15 ms

10–100 ms

Energy per Inference

1–10 µJ

30-150 µJ

100-2,000 µJ

On-Device Memory

Up to 2MB SRAM
+ 2MB MRAM

up to 1MB SRAM

2MB SRAM
with optional 2MB MRAM

Multimodal Sensor Fusion

Yes, milliwatt-scale

Limited

Limited

Always-On Viable?

Yes

No

No

Evaluation Platform

SDK Description

Seamless porting of AI models to the ECS-DoT

Customer AI
Models

EMASS AI Toolkit

Customer AI
Models

Customer
AI Models

EMASS AI Toolkit

AI models optimized
and mapped to the
ECS-DoT
DL Accelerators

Get in Touch with EMASS
Explore Edge AI Innovation