Keynote Speakers
|
(EPFL) |
Title: Design of Energy-Efficient and Reliable Wearable Systems for the Internet-of-Things Era Abstract: The evolution of semiconductor process technologies has enabled the design of low cost, compact and high-performance wearable embedded systems. These new wearable systems are computing platforms with multi-processor system-on-chip (MPSoC) architectures that can be deployed ubiquitously, and can potentially provide computing capacity for real-time analysis of human bio-signals to continue the progress of our society in the new era of Internet-of-Things (IoT). However, the inherent resource-constrained nature of wearable embedded systems, coupled with the power requirements of MPSoC architectures, can result in degraded performance and unreliable behavior, or a global energy crisis if they are massively deployed in the IoT era. Therefore, in this talk I will show the impact of potential hardware misbehavior induced by reliability issues and scaled voltages in MPSoC wearables nodes. Then, I will advocate that the inherent resilience of latest algorithms in wearable bio-signals monitoring, as well as a better understanding how living organisms operate, can allow us to conceive new cross-layer design paradigms for the next-generation of energy-efficient and reliable wearable systems in the IoT era. This new approach exploits more on-board intelligence, hardware specialization and the selective application of different robust techniques to gracefully scale the energy consumption of wearable MPSoC architectures according to the required output quality for the target biomedical application. |
|
(Microsoft Research) |
Title: Stamping Out Concurrency Bugs Abstract: The shift to multi-core architectures in the past ten years pushed developers to write concurrent software to leverage hardware parallelism. The transition to multi-core hardware happened at a more rapid pace than the evolution of associated programming techniques and tools, which made it difficult to write concurrent programs that are both efficient and correct. Failures due to concurrency bugs are often hard to reproduce and fix, and can cause significant losses. In this talk, I will first give an overview of the techniques we developed for the detection, root cause diagnosis, and classification of concurrency bugs. Then, I will discuss how the techniques we developed have been adopted at Microsoft and Intel. I will then discuss in detail Gist, a technique for the root cause diagnosis of failures. Gist uses hybrid static-dynamic program analysis and gathers information from real user executions to isolate root causes of failures. Gist is highly accurate and efficient, even for failures that rarely occur in production. Finally, I will close by describing some ongoing work we are doing to solve key reliability and performance problems of emerging software systems. |
|
(ETH Zurich) |
Title: The RowHammer Problem and Other Issues We May Face as Memory Becomes Denser Abstract: We will discuss the RowHammer problem in DRAM and how it poses a new system-wide security vulnerability. RowHammer is the phenomenon that repeatedly accessing a row in a modern DRAM chip causes errors in physically-adjacent rows. It is caused by a hardware failure mechanism called read disturb errors. The Google Zero Project recently demonstrated that this hardware phenomenon can be exploited by user-level programs to gain kernel privileges. Several other recent works work demonstrated other attacks exploiting RowHammer, including remote takeover of a server vulnerable to RowHammer. We will analyze the root causes of the problem and examine solution directions. We will also discuss what other problems may be lurking in DRAM and other types of memories, e.g., NAND flash and Phase Change Memory, which can potentially threaten the foundations of reliable and secure systems, as the memory technologies scale to higher densities. |
|
(TU Graz) |
Title: Dependable Internet of Things Abstract: Wireless networked embedded systems are increasingly used for safety-critical applications, where even under harsh environmental conditions dependability requirements must be met. In this talk we introduce the Dependable Things research center at TU Graz and present recent results on improving the dependability of wireless communication and localization, embedded computing, and networked control for the Internet of Things. |
|
(TU Darmstadt) |
Title: Everything You Code Can and Will be Re-used Against You: On the Challenges of Mitigating Code-Reuse Exploits Abstract: Memory corruption and memory disclosure vulnerabilities are still a persistent source of threats against software systems, although known for over two decades. The main problem is that modern software still contains vast amount of unsafe, legacy code. Moreover, exploitation techniques are rapidly evolving and often incorporate increasingly sophisticated techniques, which can be used to bypass all widely deployed countermeasures such as Data Execution Prevention (DEP) or Address Space Layout Randomization (ASLR). This has recently motivated many researchers in academia and industry to make considerable efforts on improving defenses against modern code-reuse exploits. It seems that there is a strong desire in our community to build secure systems from unsafe code! Hence, many software-hardening solutions have been proposed, some of which are based on hardware support. Recently Intel has released new specification on Control-Flow Enforcement Technology (CET) for x86/x64 to mitigate code-reuse techniques. However, even though these solutions significantly raise the bar for exploitation, new attacks are continually discovered, and no ultimate solution seems to be in sight. This talk gives an overview of the continuing arms race between code-reuse attacks and mitigation techniques and their nuances, particularly the hardware-based defenses. We then highlight and discuss the effectiveness and usefulness of recent approaches. The game is not over yet. |
