Like-for-like, across clouds
Compare
A list price is not a measurement until you say what it measures. Every number here is normalised onto one stated unit before anything is compared — and every number shows the SKUs it was built from, so you can check the working rather than take it on faith.
Why these prices can be compared at all
Vendors do not sell the same shape. Google prices a machine's cores and its memory separately; AWS prices the whole instance. So a like-for-like figure is composed — four cores plus sixteen gigabytes on one side, the instance type whose spec is four cores and sixteen gigabytes on the other. Where both vendors publish the same unit already, such as a gigabyte-month of object storage, the comparison is direct. Every result is scoped to one region group and one pricing model, so a committed rate is never held up against an on-demand one.
What are you buying?
Compute
Databases
Storage
Networking
Not enough to compare here
Only Google Cloud publishes something that qualifies for this combination at 4 vCPU · 16 GB in APAC on on-demand pricing. A one-sided chart is not a comparison, so there isn't one.
How each number was built
The cheapest qualifying configuration per cloud, and the SKUs it is made of
- GCPC2D series$0.1199per hour
Components of the gcp price 4 × C2D AMD Instance Core running in Mumbai unchanged — $0.0195/h 16 × C2D AMD Instance Ram running in Mumbai unchanged — $0.002613/GiBy.h Cheapest of 7 configurations that qualified · median $0.1889 per hour
No qualifying configuration from AWS under these pivots. Absent, not zero.
Where it is cheapest
Same configuration on on-demand pricing, by geography
| Region | Google Cloud |
|---|---|
| United States & Canada | $0.1375 |
| Europe | $0.1444 |
| Asia Pacific | $0.1199 |
| South America | $0.1966 |
| Middle East & Africa | $0.1513 |