How can I use 96 CPUs with Compute Engine Google Cloud?

Admin 1
Viewed 1085

I created my VM instance in Google cloud with no problems: 2 CPUs and 8GB RAM. However, I would like to increase it to 96CPU's and 138GB RAM. However, I am receiving the error message I can't have more than 24 CPU is my area (I tried all the areas and all say the same). On top of that, I also tried to create 24 CPU's in 2 different areas and I received the message that my maximum global CPU's is only 32 :-(

How can I fully use the power of more CPU's.

PS: I am using Windows Server 2019 as OS.

TIA

google-compute-engine google-cloud-platform cpu
2 Answers

There is a cores quota. You might want to increase it.

CPU quota refers to the total number of virtual CPUs across all of your VM instances in a region. CPU quotas apply to running instances and instance reservations. Both normal and preemptible instances consume this quota.

To protect Compute Engine systems and other users, some new accounts and projects also have a global CPUs (All Regions) quota that applies to all regions and is measured as a sum of all your vCPUs in all regions.

For example, if you have 48 vCPUs remaining in a single region such as us-central1 but only 32 vCPUs remaining for the CPUs (All Regions) quota, you can launch only 32 vCPUs in the us-central1 region, even though there is remaining quota in the region. This is because you will reach the CPU (All Regions) quota and need to delete existing instances before you can launch new instances.

From: https://cloud.google.com/compute/quotas#cpus

To increase cores quota refer to: https://support.google.com/cloud/answer/6376374?hl=en

user_4010250

There's no affordable cloud hardware that has such specs, and Google wouldn't be providing such anyway. I guess if you found some "cloud" provider with modern big iron NUMA, you'd get that, but still it would be NUMA so shuttling data between the CPUs would incur overhead, and you'd be paying out the nose. You'll simply want to run multiple instances of that VM: since you can clearly parallelize the work to use multiple CPUs, then the process that has all those threads can be easily split to work across as many VMs as needed.

user_13296520
Related