My impression is that Google has a first class general purpose instance but you don't really get the breadth of options that EC2 will give you.
You bring up Local SSD. Google's Local SSD is just badass by comparison:
- 680,000 Read and 360,000 Write IOPS included in the cost [0]
- $0.218 per GB per month. Instance cost is separate.
- Again, you can attach these to any instance type (hence the point on fragmentation of instances on EC2)
- AWS goes up to 365,000 Read and 315,000 "First Write" IOPS. Only if you buy an i2.8xlarge [2]
- An i2.8xlarge is $6.82 per hour.
You do the math :)
And someone else did more comparisons here [1]
[0] https://cloud.google.com/compute/docs/disks/performance
[1] https://medium.com/google-cloud/new-google-cloud-ssds-have-a...
[2] http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/i2-instan...
Followed by waiting on GPUs and other user accessible accelerators of course.
1) The highest memory instance on Google is about 200GB, on AWS it's 2TB.
2) Google doesn't have GPU instances yet. They're been announced for 2017.
Other than that, google instances have more options and more flexibility.
Which is why I find it so annoying when I see these price comparison articles that don't really use the SpotFleet API correctly.
Just in the "only instances" space you mention Google has a different strategy. A couple things to mention:
- Google doesn't have a broad spectrum of instance types. No storage-optimized or networking optimized. Instead, any instance can just get great storage attached to it, and any instance gets best-in-class networking.
- Google's Preemptible VMs are like Spot instances but fixed 75-80% off. Again, with less fragmentation against instance types + fixed cost, much easier to rely on Preemptible VMs imo.
- Google's Load Balancer is global, scalable, anycast IP driven, and backed by Google Network. If your packets originate in, say, New Zealand, they'll be talking to a "GCLB instance in our pop in Sydney", which will carry packets on Google's backbone to the VMs.
- Custom VM sizes - you can set your own VM/RAM combination for instances.
- Live Migration. Google manages instance health and maintenance for you, without forcing restarts.
(Work at Google but not on GCE.. and I don't get paid to post here)