two important facts about storage at scale: 1) google is (to some degree) disk throughput bound which is why they charge for it 2) google runs almost entirely (on a byte basis) on HDDs

Posters: cis_female. cis_female

2 likes recorded by the source corpus.