2 articles tagged with #gpu.
A practical guide to genomic foundation models for diagnostic DNA and RNA testing. Eight clinical tasks, which ones a foundation model has been tested on, what to use where it has not, which model to pick when you do want one, and what it costs to run.
GPU and DRAGEN pipelines cut a whole-genome run from 17 to 39 hours on CPUs to between half an hour and about two hours, and the one study that compared their calls with the CPU pipeline found close agreement. No study reruns a pipeline to test reproducibility, and FP8 gave a genomic foundation model at most 19.3% more throughput.
