Yes, because as we all know, university teams don't have access to hardware, but are swimming in money. (obviously the whole point of doing cloud for the cloud providers is that owning the hardware is cheaper)
Whilst I would agree that university teams probably should use the resources the cloud providers make available freely, they should probably stay away from actually using capacity on the cloud and instead have their own hardware.
Besides, what I keep hearing from machine learning researchers is that no matter where you work, developing on your own machine ... there's no beating that, time and productivity wise.
I used to do IT for a large research university. I can confirm that we were practically drowning in funding for new hardware. If we made a business case for upgrading our infrastructure it wasn't uncommon to get an extra million or so dollars in the budget without much thought.
Can't disagree with that, but for some genomics work it's just not a realistic option when you can get output from a slice of mainframe in 1/10th the time, and that time is like 2 and some change days.
