2018 IEEE/ACM International Symposium on Microarchitecture (Micro-51)

The second keynote was delivered by Ruby B. Lee, the Forest G. Hamrick professor in engineering and a professor of electrical engineering at Princeton University.

Lee spoke about “Security Aware Microarchitecture Design.” At her Princeton Architecture Lab for Multimedia and Security (PALMS), Lee ‘s research includes designing security-aware architectures for processors, smartphones and cloud computing. Prior to Princeton, Lee was chief architect at Hewlett Packard computer systems.

“Recent covert and side channel attacks, like Meltdown and Spectre, have alerted the world to the security dangers of attacks on hardware architecture features. Even though microarchitecture performance features, like speculation and branch prediction, may be implemented correctly according to current computer architecture specifications, they can be exploited to leak security-critical or privacy-sensitive information to unauthorized entities. How then, can we design microarchitecture that improves performance without degrading security? Can we design architecture that improves security and performance at the same time?” she wrote in a summary of her remarks.

“The good news is that this ‘era of security’provides both critical challenges and exciting opportunities for new design principles and innovative strategies for joint performance-security optimization, security-functional-timing verification and new tools,” she said.

Mike Davies, who is director of Intel’s Neuromorphic Computing Lab, addressed “Neuromorphic Principles for Efficient Self-Learning Microarchitecture” in the keynote address on the third and final day of programming. The conference was held October 20-24.

“Earlier this year, Intel published its Loihi neuromorphic research chip in IEEE Micro (magazine). This novel processor implements a microcode-programmable learning architecture supporting a wide range of neuroplasticity mechanisms under study at the forefront of computational neuroscience. By applying many of the fundamental principles of neural computation found in nature, Loihi promises to provide highly efficient and scalable learning performance for supervised, unsupervised, reinforcement-based, and one-shot paradigms. This talk describes these principles as applied to the Loihi architecture and shares our preliminary results towards the vision of low power, real-time on-chip learning,” Davies said in a summary of his remarks.