I'll amend it to calling a hyperthread a vcpu.
People are constantly confused about this. It's not just me. There's very little common terminology across leaders in this space. Think VMware, Digital Ocean, Bluemix, Tier 1 cloud providers, in house data centers, etc. Some talk real cores, others talk hyperthreads. Business and non techie IT people do their own math without knowing the 2X difference.
If you put (hyperthreads) next to "vcpus", when appropriate, it would be more clear.
"Direct connect pricing for egress traffic ("Data Transfer Out") is here"
Attend an AWS sales meeting with a F500 company. It's never mentioned, even when talking about obvious problem areas...like moving a data warehouse to AWS, while leaving the ETL and analytic tools on premisis. People like me are left holding the bad news bag.
AWS sales people talk up the "all in" business case where all your stuff is in the cloud. They are suspiciously quiet about the transition period where the 10 to 100x market bandwidth pricing of AWS kills you. That's especially frustrating for AWS direct connect where your global network adds no value. In my mind, AWS direct connect should be competitive with market rates for bandwidth. I understand the premium for general internet connectivity because your global presence and investments add value. For direct connect, it's just plain old gouging. You're penalizing me for not going all-in on day one, which would be risky and stupid.