Company: Confidential
Job Title: Postdoctoral Position at the Intersection of AI and Cloud Computing
Job Number: 86203
Location: Nashville, Tennessee, US
Job Description
The Simulation, Optimization, and Learning (SOL) Laboratory in the Department of Computer Science at Vanderbilt University seeks to immediately fill a Postdoctoral Researcher position for a research project on using AI/deep learning to optimize cloud computing resource usage. The postdoctoral scholar, supervised by Professor Hyde (https://dabh.io), will have the opportunity to get involved in the development of theoretical foundations, development of software tools, and systems/experimental research using real-life problems and datasets.
The specific problem we aim to solve is to make it easier for users of cloud computing to select the compute instance type (e.g., AWS EC2 instance type) that is best suited for a given workload they are trying to run. (There are many instance types available, spanning e.g. CPU and GPU instances, and more instance types frequently become available.) For instance, if a user has just run a particular Python script with an 8-GPU node, it is likely that if we encounter that same Python script again, the user is likely to want to use an 8-GPU node again. In another case, we may have a precompiled code (e.g., from a C++ codebase), and we want to use static/binary analysis to determine whether any parallelism or GPU code/libraries are included or linked; we could then determine whether to exclude GPU or many-core instances from consideration (this doesn’t necessarily require AI). Moreover, we’ll want to instrument cloud computing nodes when a job is being run—monitor CPU, RAM, GPU, and disk usage, at the very least—to determine if nodes are being underutilized (in which case, less powerful instance types could be selected) or overutilized (e.g., a job crashes because the node runs out of RAM). This high-dimensional data can be used in a neural network model to further help make an automatic recommendation of the instance type to use for a given job.
We are seeking motivated, self-driven postdoctoral researchers with a solid technical background and experience in areas related to cloud computing, AI, and possibly static analysis. We give strong priority to candidates with real-world experience in software engineering, particularly those who have built cloud computing platforms and who have expertise in designing and training neural networks. Skills with Python, Node.js, AWS, GCP, Azure, Datadog, Pytorch, git, etc. are highly valued. A Ph.D. in Computer Science or a closely related field is required.
Application Deadline: 2025-03-31
To apply for this job please visit jobrxiv.org.