Summary
We live in an era of unprecedented data explosion, where autonomous intelligent systems constantly generate and process enormous amounts of data in real-time. For example, today a single autonomous vehicle can generate nearly 40 terabytes of sensor data per hour, which is equivalent to streaming 6,000 Netflix movies simultaneously. Interpreting such massive data with artificial intelligence (AI) requires computing hardware that is both extremely fast and highly energy efficient while providing large storage capability. However, today’s computer chips face fundamental limitations. Most chips ar
