Projects - 2021


Mahlet Dereje Haile

Project title: Quantifying Bird Skeletons
Faculty Mentor: David Fouhey

Description: Our lab is broadly focused on building 3D representations of the world and understanding human/object interaction. Projects include learning about: navigating environments, object articulations, commonsense physical properties of objects, and hand grasps.

Mahlet Dereje_FINAL_9-30-21.mp4

Tibebu Getachew Wassie

Project title: 3D Reconstructions from Partial Observations
Faculty Mentor: David Fouhey

Description: Our lab is broadly focused on building 3D representations of the world and understanding human/object interaction. Projects include learning about: navigating environments, object articulations, commonsense physical properties of objects, and hand grasps.

Tibetu Wassie_FINAL_9-28-21.mp4

Adona Tesfaye and Samuael Adnew Birhane

Project title: Proxy-Web
Faculty Mentor: Baris Kasikci

Description: Currently, there is no benchmark suite representative of real-world Web services. Existing suites like CloudSuite do not exhibit the typical challenges faced by modern Web applications (e.g., high-degree of frontend stalls due to misses in the instruction cache and the branch target buffer, an order of magnitude higher branch mispredictions, etc.). Similarly, frameworks like DaCappo are Java-centric and are not relevant to big Web services companies like Facebook and Google, where most of their backend code is written in C++ for performance reasons. Getting access to the real code that is being run at these companies is usually not possible, but there are numerous public works detailing the characteristics of these applications.

The idea behind this project is to both (a) mine publicly available data on Web app characteristics to create proxy applications that are representative of the real workloads and (b) consult Web service companies to determine how representative the suite is. Insights gained from developing this suite will help drive the development of accelerators in this space as well as architectural and microarchitectural support. A broader goal of the Proxy-Web project is to help to make potential projections about where Web service applications are headed in the future.

Adona Tesfaye_Samuael Adnew_FINAL_9-30-21.mp4

Yabsra Abdurahman

Project title: Hazel Notebooks: Building a Better Jupyter
Faculty Mentor: Cyrus Omar

Description: The popular Jupyter lab notebook environment is powerful, but it has a problem: results stored in a notebook are not reproducible, because the user can execute cells out of order. In our group, we are developing a new live functional programming environment called Hazel ( Right now, Hazel does not support multiple program cells and other notebook features. This project will turn Hazel into a next-generation version of Jupyter by adding support for notebooks with multiple cells and multiple pages, with dependencies between them. We will solve the reproducibility problem by developing a mechanism conjectured in a recent paper in our group: fill-and-resume.

Yasbra Abdulrahman_FINAL_9-28-21.mp4

Tigist Shiferaw and Bereket Ngussie

Project title: Verified implementation of a quadcopter controller
Faculty Mentor: Jean-Baptiste Jeannin

Description: This project is about using a new programming language with formal verification capabilities to program a quadcopter. The language is designed to be used for drones and cars, and is developed in our lab. The student will use this new language to program basic control capabilities of a quadcopter drone, then include some collision avoidance capabilities. The quadcopter will be at Michigan but there will be opportunities to run tests remotely.

Tigist Shiferaw_Bereket Bereket Ngussi_FINAL_9-28-21.mp4

Miruts Hadush and Tersiteab Mersha Adem

Project title: Novel Architectures to Compute with Graphs
Faculty Mentor: Valeria Bertacco

Description: More and more applications rely on graphs as the underlying data structure: from social networks, to internet’s web connections, to geo maps, to ML algorithms and even consumers’ product preferences. The performance of these algorithms is often limited by the latency of accessing vertices in memory, whose access present poor spatial locality. The goal of this project is to boost the performance of graph-based algorithms by developing hardware and software solutions to this end: we plan to work on the data layout, on ad-hoc data structures and on designing dedicated hardware acceleration blocks. We hope to boost the performance of graph traversals by 3-5x.

Miruts Hadush_Tersiteab Mersha_Accelerator_FINAL_9-28-21.mp4

Miruts Hadush and Tersiteab Mersha Adem

Project title: From high-level language to hardware — without the hardware design
Faculty Mentor: Valeria Bertacco

Description: This project explores a new hardware design flow, where the starting point is an application specified in a domain-specific language (more specialized than C) like Halide or GraphIt, and the endpoint is a hardware system equipped with specialized hardware accelerators, so to execute the application much faster than it would be possible in software. To reach the endpoint, we will work on the back-end of the compiler, so to target the primitives available in the hardware accelerators.

Miruts Hadush_Tersiteab Mersha_Benchmark Suite_FINAL_9-30-21.mp4

Galane Basha Namomsa and Plato Gebremedhin

Project title: Security Project: Stochastic Side Channel Attacks
Faculty Mentor: Todd Austin

Description: Side channel attacks are one of the primary means to attack modern hardware. The attacks use visible properties of a system, like timing, power, and electromagnetics, to infer information about the secrets held within. Eliminating these vulnerabilities is primary focus of hardware security research today, and the leading approach to stop these attacks is through randomization techniques. For example, to stop cache timing side channels, CPU developers are starting to deploy cache mapping randomization techniques. The goal of this project is to demonstrate that randomization techniques are a futile direction to pursue in stopping side channel attacks. We will do this by developing stochastic side channel attacks, which utilize randomized probes to infer secrets, thus, they are immune to all randomization defenses! Along with these advanced attacks, we will start to work on more durable defenses for hardware side channels.

Galane Basha_Plato Gebremedhin_FINAL_9-28-21.mp4


Bereket Barma and Amanuel Solomon

Project title: Using Electricity Smart Meter Data to Understand Opportunities for Residential Energy Efficiency and Demand Response
Faculty Mentor: Johanna Mathieu

Description: The student researcher will support a new research project: "Reducing Barriers to Residential Energy Security through an Integrated Case-management, Data-driven, Community-based approach" funded by the US National Science Foundation. See: . The project will leverage electricity smart meter data from households in Detroit, Michigan to develop algorithms to pair electricity consumers with energy efficiency and demand response programs. The student researcher will contribute to a literature review of approaches to analyze household smart meter data and also implement a variety of existing analysis techniques, possibly including clustering and disaggregation, to characterize and interpret the data and look for opportunities for energy savings and demand flexibility. The student researcher will also explore the design of novel algorithms.

Amanuel Solomon_Bereket Barma_FINAL_9-28-21.mp4