According to a report by Technology New Report, scientists at the University of Pennsylvania in the United States have successfully developed a memory device capable of operating continuously for 60 hours at high temperatures up to 600°C. This breakthrough in temperature resistance is more than double that of current commercial storage devices, demonstrating the outstanding reliability and stability of the new memory device. It is expected to play a significant role in extreme environments that could cause electronic or storage device failures and provide support for artificial intelligence systems conducting intense computations under harsh conditions.

Researchers note that this memory device is a non-volatile type, capable of retaining stored information over the long term even when no power is supplied. In contrast, traditional silicon-based flash memory begins to fail when temperatures exceed 200°C, leading to device malfunction and data loss.

The new memory device is made using a material called ferroelectric aluminum scandium nitride (AlScN). AlScN offers significant advantages in storage because it can maintain specific electrical states, such as on and off, at higher temperatures after removing an external electric field. Its unique crystal structure results in more stable and robust atomic bonds, providing not only heat resistance but also durability. Additionally, the design and performance of the storage device support rapid switching between different electrical states, which is crucial for high-speed data reading and writing.

The memory device has a metal-insulator-metal structure, consisting of nickel and platinum electrodes and a 45-nanometer-thick layer of AlScN. This structural design enables the memory device to be compatible with high-temperature silicon carbide logic devices and work in conjunction with high-performance computing systems designed for extreme temperatures.

Researchers state that this new type of memory device is a “memory-enhanced computing” device with high stability, allowing for tighter integration of memory and processing elements, thereby increasing the speed, complexity, and efficiency of computations. They will continue to explore the application of the new device in AI systems operating in extreme environments.