06-20-2016, 05:56 PM
Now one question remains: why sacrifice 10% performance for the offered flexibility with 100% uptime guarantee? You can have an uptime of 99.9% without the effort of virtualizing. (unless you need 1h maintenance per day to update the hardware…)
Agility for hot-add/remove ram/cpu? You never need to change that for a GPU cluster.
Availability: yeah. Why the hell do I need 100% availability for a hash cracking cluster? What is the gain? If it's 1% better uptime (doubtful) that still does not account for the 10% of performance loss.
Agility for hot-add/remove ram/cpu? You never need to change that for a GPU cluster.
Availability: yeah. Why the hell do I need 100% availability for a hash cracking cluster? What is the gain? If it's 1% better uptime (doubtful) that still does not account for the 10% of performance loss.