(Note the GKE SLA is for regional clusters, which is what you should be doing if you care about uptime. The zonal cluster SLA is 2.5 nines. I couldn't find a difference in EKS, maybe there's an equivalent better SLA for regional clusters I couldn't find.)
So, per my original comment, I am surprised. (Having never used EKS directly I have no idea what their actual uptime is; in my experience GKE has been way higher than 3.5 nines, but obviously I don't have enough data to make statistically significant observations on this.)
I'd be surprised if they have the same operational maturity, which translates to uptime, which was what I was talking about.
A quick search turns up these SLAs:
<bad link>
vs
https://cloud.google.com/blog/products/containers-kubernetes...
EKS: 3 nines
GKE: 3.5 nines
So Google is committing to 1/2 the downtime.
(Note the GKE SLA is for regional clusters, which is what you should be doing if you care about uptime. The zonal cluster SLA is 2.5 nines. I couldn't find a difference in EKS, maybe there's an equivalent better SLA for regional clusters I couldn't find.)
Edit - that EKS SLA link formatted weirdly, and so I did a little more digging and found a more recent SLA which matches GKE: https://aws.amazon.com/about-aws/whats-new/2020/03/amazon-ek...
So, per my original comment, I am surprised. (Having never used EKS directly I have no idea what their actual uptime is; in my experience GKE has been way higher than 3.5 nines, but obviously I don't have enough data to make statistically significant observations on this.)