News & Events

Read the latest news about Aspinity and find out where you can hear more

AnalogML

Myth #5: AnalogML™ is just for machine learning applications

Myth #5: AnalogML™ is just for machine learning applications

AnalogML™ is not just an ultra low-power analog neural network, but has other sophisticated analog processing functionality that together reduce power, size, and cost in always-on endpoints. Read more

Myth #4: CMOS Process Variations Make AnalogML™ Unpredictable

Myth #4: CMOS Process Variations Make AnalogML™ Unpredictable

How does analogML™ overcome the repeatability challenges associated with analog process variation? Read our latest installment of our 5 Myths about AnalogMl™ series to find out. Read more

Myth #3: AnalogML™ extends battery life at the expense of wake word accuracy

Myth #3: AnalogML™ extends battery life at the expense of wake word accuracy

How do newer, lower-power system architectures that rely on analog sound data to activate the WWE manage preroll to maintain wake word accuracy? They can't - EXCEPT for those using the analogML core. Read more

Strategies for power efficient machine learning at the edge

Strategies for power efficient machine learning at the edge

Power-efficiency at the edge is a huge challenge for always-on system designers. Read how analog is being used in several different ways to solve the problem and compare the methods to decide what is best for your application in this article by CEO Tom Doyle. Read more

Awesome Analog Artificial Neural Networks (AANNs)

Awesome Analog Artificial Neural Networks (AANNs)

Max Maxfield talks about Aspinity's analogML core and the emergence of analog as an important component of machine learning at the edge. Read more

Join our Mailing List

Join our mailing list to receive our newsletter and never miss out on the news from Aspinity.