GCP, like AWS has instances with 4 E7-8880v4 CPUs (up to 160 vCPUs) and 4TB of memory. That's a pretty significant single point of failure though, and not at a

GCP, like AWS has instances with 4 E7-8880v4 CPUs (up to 160 vCPUs) and 4TB of memory. That's a pretty significant single point of failure though, and not at all elastic. What type of application/workload prevents you from scaling horizontally?