Comment by groundthrower

Okay, nice. And if I would like like 32 vCpus? Having an application today that has a huge degree of parallelism, but utilizing an external cloud provider that offers dedicated machines with very affordable pricing. Would really like to use lambdas instead though.