Sylvain Dubois, Vice President of Business Development and Marketing at Crossbar, demonstrates the company's latest embedded vision technologies and products at the 2019 Embedded Vision Summit. Specifically, Dubois demonstrates how near-memory computing can significantly improve the overall performance of neural network hardware accelerators. He also showcases how resistive RAM (ReRAM), designed for parallel computing, is able to achieve billions of Object LookUps Per Second (OLUPS). He concludes with a performance benchmark of Crossbar XPU vs a quad-core Arm Cortex-A53 with DDR4, showing more than a 30X performance improvement while consuming 500X less power.
