A university is running a DNA decoding project that will take seven years if it runs on its current internal mainframe. The university negotiated a deal with a large cloud provider, which will…

Storage for results is a secondary concern - it is needed, but it does not affect the speed or feasibility of the compute processing itself.

A larger network pipe improves bandwidth but does not reduce latency, and a distant high-bandwidth link is still slower for large sequential transfers than a nearby lower-bandwidth link.

Genomic datasets are extremely large, and the round-trip time for uploading raw DNA data and downloading decoded results is directly tied to the physical distance to the datacenter. A closer datacenter reduces network latency and improves effective bandwidth utilization, which can be the primary bottleneck when transferring petabyte-scale datasets. Since compute is already being donated, geographic proximity is the most impactful additional resource to negotiate.

Compute resources are already the subject of the donation arrangement described in the question; asking for any available compute does not address the data transfer challenge inherent in a remote processing model.