The second generation SPnvSRAM family delivers high-speed non-volatility, preventing data-loss even in harsh industrial, consumer operating environments over extreme temperatures. The SPnvSRAM Low Energy series offers 1 to 32-Mb densities at extended-temp industrial-grade specifications. These devices are now available in low pin count, small package options, making them ideal for a broad range of battery-powered, low-energy industrial, consumer, wearable and IoT applications.
The second generation of low energy SPnvSRAM is offered in 20-MHz Serial Peripheral Interface (SPI) as a byte addressable memory, thus eliminating the need for software device drivers. These devices further simplify the software design by not requiring the host controller to issue sleep and wake-up commands as is common with other SPI devices. As a byte addressable memory, there is no delay or buffering, which assures the integrity of the data being written. These devices are offered in Extended Industrial temperature in a variety of small form factor Industry pin compatible RoHS packages. The family is also offered in a wide range of voltages from 1.71V to 3.6V, suitable for today's industrial IoT designs as well as future generations of low power wearables. MRAM's inherent immunity to Alpha particles also makes it an ideal solution for devices that are regularly exposed to radiation.
The higher densities (up to 32Mb) in the same small form factor enable customers to simplify designs and eliminate multiple devices that take up valuable board space. As customers' designs grow, the devices can be simply upgraded to the next density within the same family architecture.
The persistence offered by SPnvSRAM is also enabling a new generation of IoT nodes capable of Machine Learning where the inference algorithms do not have to be reloaded every time after device wakeup. These IoT nodes with Convolutional and Recurrent Neural Networks are being designed with simple binary weights that trigger coarse inference in real time and can further use the cloud for near human level inference performance based on traditional Stochastic models.