Sirumalla Awarded $5000 in Amazon Web Services Credit
Chemical Engineering MS student Krishna Sirumalla, won an award of $5000 in Amazon Web Services (AWS) credit to develop his project, “Predicting Kinetics Using Graph Convolutions on Transition State Graphs.” This three-part project will utilize AWS’s service Elastic Compute (EC2). In this project, Sirumalla is representing a transition state of chemical reaction as a graph and storing the engineered features on nodes of the graph. Additionally, he is training a neural network model on training data about GPUs. Lastly, he is using the trained model to predict kinetics. The Elastic Compute service enables one to launch a server with a number of CPUs or GPUs and use that server to run calculations. Sirumalla writes, “The $5000 in credits means I can launch servers with GPUs to train my model and also launch servers with CPUs to generate training data.”