®

Syntiant  Always-On Sensor & Speech

Recognition Processor

NDP100

NDP100 Neural Decision Processor™

Ultra-Low-Power Consumption

Micro-watt level active power consumption, while always-on and listening

Keyword Speech Interface

The Syntiant® NDP architecture is built from the ground up to run deep learning algorithms. With embedded Syntiant Core 1™, the NDP100 achieves breakthrough performance by highly coupled computation and memory, exploiting the vast inherent parallelism of deep learning and computing at only required numerical precision. The devices combine these elements to achieve approximately 100x efficiency improvement over stored-program architectures such as CPUs and DSPs. 

Compact Size

Neural Processing

Neural network with over half a million parameters providing large scale parallelism for sensor-based applications
Extremely small 1.4mmx1.8mm, 12 ball WLBGA package for space-constrained applications
Created from the simultaneous classification of dozens of keywords

Ease of Use

Multiple, flexible methods to access neural processing engine where the user has full customization of applications and post-processing

Applications

The NDP100 enables speech interfaces in the smallest systems, replacing traditional tactile interface elements (buttons, switches and dials) and supporting entirely new form factors and usage models designed to wake up to speech rather than touch. 

mobile phones • ear buds and hearing aids • bluetooth headsets • smart watches • IoT endpoints • remote controls • smart speakers ​

Key Features

+    12 ball WLBGA package for extremely space-constrained environments
 

+    Dual PDM microphone input or PCM-over-SPI input
 

+    Stereo/mono I2S interface multiplexed with PDM
 

+    Direct access to neural network over SPI for sensor applications
 

+    Frequency-domain, time-domain & batch input models
 

+    16-bit input holding tank with faster than real-time SPI extraction
 

+    Embedded Arm Cortex-M0 processor with 112KB SRAM
 

+    Integrated clock multiplier and dividers support low-frequency clock source or external clocking

+    Optimized interrupt and SPI slave interface

 

+    Onboard firmware security and authentication

 

+    English speech service for keyword training

 

+    Software Development Kit (SDK) integrates in any software environment

 

+    Training Development Kit (TDK) to enable the use of standard frameworks such as TensorFlow for customer-programmed applications

 

+    Support for 64 output classifications

 

+    Active power consumption of <140 µW while recognizing words

  • White Twitter Icon
  • White LinkedIn Icon

SYNTIANT IS A REGISTERED TRADEMARK OF SYNTIANT CORP.

IRVINE, CA