Reading this just gets me down. I wouldn't mind if these were phrased as questions but no, we get blanket assertions. > The network would be much faster then l

Reading this just gets me down. I wouldn't mind if these were phrased as questions but no, we get blanket assertions. > The network would be much faster then local hard-drives. Figures? Are you distinguishing between bandwidth and latency? (no you aren't). How long are reads/writes going to take compared to a local disc: HDD seek times ~ 10 milliseconds, SDD in the microseconds. Speed over a network to 'the cloud' - tens of milliseconds (try pinging BBC from the UK, ~22 milliseconds. Example.com is 98 MS). > 10Gbe is faster then most hard drives. is 1 Gig/sec, approx. A couple of SSDs will match that in theory, in practice I doubt you'll get anything like saturation from a distant datacentre > With a Tbit network it would be many times faster to load from the network then to load from the computer's own memory OMG. C? > and everything would be downloaded, searched and discarded if not matched OMG redux. I'm no network guy but do you think tying up dozens of routers with multi-gigabytes of data you're going to discard, is free? efficient in energy or time? Some reading matter. <https://dzone.com/articles/latency-numbers-everyone-should-know> https://dzone.com/articles/latency-numbers-everyone-should-k... <https://www.cloudways.com/blog/how-server-location-affects-latency/> https://www.cloudways.com/blog/how-server-location-affects-l... That kind of response would be OK if it asked questions, expressed a knowledge of the limits of their knowledge, I'd be ok with that and pointing the asker to relevant material but as it stands, as in that famous quote, it isn't even wrong.