NeuroMem

Overview
Neuromorphic Computing?
Product stageSegments
Minimum Viable Product
?
Hardware - Circuit
?

NeuroMem is a neuromorphic technology company that develops brain-inspired chips and neural networks for pattern recognition and machine learning applications. The company's flagship product is the NM500 chip, a digital neural network featuring 576 neurons capable of intrinsic learning and recognition of patterns derived from various data types, including images, sounds, instruments, and text. NeuroMem's technology employs in-situ learning and auto-correction mechanisms, allowing neurons to learn new examples without supervision and make decisions collectively.

The NeuroMem neurons operate using a Winner-Takes-All parallel logic, enabling rapid processing within microseconds. Unlike conventional deep learning approaches, NeuroMem's technology does not require access to massive databases of annotated models for training. The neurons can be pre-trained by loading knowledge files previously built by a NeuroMem network. The company's neural networks provide real values instead of just probabilities, acknowledging uncertainties in classifications when appropriate.

NeuroMem offers various hardware solutions, including the NeuroTile, a small sensory hub for real-time machine learning, and the NeuroStamp, a powerful learning and inference engine. The company also provides the NeuroShield board, which features 576 neurons expandable to 4032, and can be integrated with Xilinx ZYNQ platforms. NeuroMem's technology is available as an IP core for FPGAs from Xilinx, Altera, and Lattice, and is also available for licensing.

To support its hardware offerings, NeuroMem provides software tools such as the NeuroMem Knowledge Builder, which allows users to train and test neurons on datasets while producing detailed diagnostics. The company also offers APIs and tools for generic pattern recognition and image recognition, integrating a cycle-accurate simulation of 8000 neurons.

Key customers and partnerships

Pisces Fish Machinery Inc. has developed and sold over 50 smart cameras powered by NeuroMem neurons to inspect fishes on filleting lines aboard fishing vessels in the North Sea. In November 2021, Neurotechnologijos installed a NeuroMem-powered monitoring system in a steel blaster furnace in Magnitogorsk, Russia, using NT Adaptive Controllers. General Vision, the company behind NeuroMem technology, has partnered with nepes in Korea for the manufacturing of the NM500 chip. Additionally, NeuroMem has been integrated into Xilinx ZYNQ platforms, including the Digilent Arty Z7 and the Avnet MiniZed, expanding its potential applications in embedded systems and edge computing.


Sources

Disclaimer: This company profile has been generated using data obtained through automated web searches and advanced generative AI technology. While we strive to ensure the accuracy and reliability of our sources, auto-generated information could be outdated or inaccurate and should be verified independently.
HQ location:
Geneva CHE
Founded year:
2014
Employees:
1-10
IPO status:
Private
Total funding:
USD 100.0 k
Last Funding:
USD 100.0 k (Equity Crowdfunding; Oct 2014)
Last valuation:
-
Key competitors
Filter by the segments to which the disruptor belongs
All Segmentsexpand
 
Loading...
Loading...
Loading...
Loading...
Product Overview
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
Product Metrics
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
Company profile
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
-
Loading...
Loading...
Loading...
Loading...
Funding data are powered by Crunchbase
arrow
menuarrow
Click here to learn more
Get a demo

By using this site, you agree to allow SPEEDA Edge and our partners to use cookies for analytics and personalization. Visit our privacy policy for more information about our data collection practices.