[zackbloom]

We have a post which compares CPU intensive workloads which should be ready after the American holiday. The summary is a 128MB lambda provides you with roughly 1/8 of a CPU core, which is therefore 8x slower than a worker.