The BigOnc project is a collaboration between SDSC and the Scripps Translational Science Institute (STSI). The concept of the BigOnc project is to create a Big Data framework that stores massive amounts of data describing the genotype, treatment regimes, and clinical outcomes for large numbers of cancer patients. The goal of BigOnc project is to allow clinicians to submit queries that consist of their patient tumor genomes, and receive information that will help them choose appropriate therapies for their specific patient/tumor. The project requires the ability to store, access, and query big data rapidly, since each uncompressed human genome approaches 1 TB. The unique architecture of Gordon will be extremely useful in evaluating the viability of the BigOnc framework, because the combination of high memory and fast data access via SSDs will allow us to analyze the large amounts of data required in an environment designed for exactly this kind of use case. The startup allocation requested here will allow us to begin benchmarking experiments so we can understand how to organize and query these large volumes of data on the Gordon architecture. Initially experiments will focus on evaluating genetic variations of patients in much smaller files (10 to 20 GB each). These preliminary experiments will help us benchmark the performance of Gordon for as a platform for this type of large data querying, and provide proof of concept for the BigOnc project. The results will guide us in scaling up to the full production version of BigOnc resource that we envision.