Unveiling the Future: A Recap of NXP Technology Days in Boston

Imagine being in a space where innovators, creators, and technology aficionados converge to push the boundaries of what's possible in the world of Edge AI. This whirlwind experience was exactly what we encountered at the recent NXP Technology Days in Boston, Ma.

Artificial Intelligence took center stage at the event, captivating attendees with in-depth discussions, demonstrations, and insights that showcased the emerging possibilities of artificial intelligence on edge devices. In addition the numerous informative sessions on the subject, the General Session led by Ali O. Ors, Global Director and AI ML Strategist, delved into the benefits of deploying AI models to the edge. A notable example highlighted during the session was the capability to train a device about a user manual, such as for a car or a washing machine, enabling it to independently respond to inquiries about the device without external assistance. This not only enhances security but also significantly boosts response times, enabling the use of these models in real-time applications.

The current challenges of edge AI is reduced accuracy and response time due to limited device storage and computational power. To compensate, quantization and pruning help reduce model size, while specialized hardware improves model efficiency. NXP's latest products feature a Neural Processing Unit (NPU) for faster speeds, up to eleven times faster than without the NPU. This can all be utilized by using NXP's eIQ Toolkit, which drastically reduces the development time of your edge application. One session demonstrated selecting, training, validating, testing, and deploying a model for identifying flowers with 60% accuracy in five minutes. The accuracy here was not the sticking point as this was sure to increase with more training and data, but the speed of development here is a major accomplishment.

Other topics of discussion at the event focused around wireless communications, NXP's support for the Linux Foundation's Zephyr RTOS, I3C, and NXP's latest NAFE family. I found myself particularly interested in NXP's Zephyr support. They currently support their FDRM development boards for the MCX micro-controller family, notably the MCXN947 NPU enabled micro-controller enabling AI on such a small platform, and the IMX8 and IMX9 NPU enabled SoCs with real-time M-core processors. I would not be surprised to see support for their newly announced IMXRT700 crossover MCU with an NPU in the near future.

Amidst the sea of new faces, there was a heartwarming reunion for myself who unexpectedly ran into an old coworker. The chance encounter served as a reminder of the interconnectedness of the tech community and the lasting impact of past collaborations. It's moments like these that make events like NXP Technology Days more than just a gathering—it's a shared journey of growth and connection.

In conclusion, NXP Technology Days in Boston was a testament to the relentless pursuit of innovation and the boundless potential of Edge AI. It was a melting pot of ideas, expertise, and creativity—a convergence of minds working towards a future where technology knows no limits.

Let's continue to push the boundaries, foster connections, and pave the way for a future where technology and imagination intertwine to create remarkable possibilities. The journey has just begun, and the future is waiting to be unveiled.

Brenden Tisler