The Industry is Hearing POLYN on AI, ML, and the Edge
We just returned from Hardware Pioneers, the third of three important industry events that POLYN took part in over the past few weeks, the others being the Design Automation Conference and TinyML EMEA Innovation Forum.
Hardware Pioneers packed a lot of people into a crowded, high-energy one-day event in London. It offered an opportunity to meet managers and engineers, industry veterans, and students. All of them are interested in new technology and in understanding the newest trends, especially anything related to AI and the edge.
It is clear that AI at the edge is all about performance and efficiency, energy saving, bandwidth allocation, and latency reduction. What we saw at all of these events is that in comparison to a year ago, more people are getting specific about these topics, asking precise rather than general questions. There is a broader understanding that operating at the edge could involve micro, milli, or watts of power depending on what you want to do.
At the TinyML event, reflecting the fast-growing community focusing on machine learning technologies and applications for thin edge computing, there was no doubt the ML sensor is emerging as a new paradigm for the future of embedded machine learning applications. It addresses such challenges as complex system integration, power consumption and latency improvement, data privacy, and security.
POLYN’s Neuromorphic Front End (NFE) provides such a solution, adding intelligence on sensor edge devices to combat these challenges.
Between the three events, we met with more people that we can count, fielding questions about our plans, our roadmap, and what types of applications we might collaborate on.
CEO Aleksandr Timofeev spoke at each of these events, and his presentations generated many questions about the technology and applications. The discussions we have had are helping us clearly separate NFE from the classification, interpretation, and decision tasks that are performed well by digital processors and clarify NASP’s role in distributed computing. We are excited that people are asking so many questions about the analog domain in trying to understand how an “old” technology can be so cutting-edge.