Hope this is the right place for this.  Apologies up front if i'm using the wrong lingo, misunderstanding concepts etc.  Please correct me.

Question about NUMA and performance.

I have BL460 gen9 with 2x 10-core Xeons and currently 160GB RAM.

However - the workload we now want to run on these hosts consists of 8-core VMs running Citrix Xenapp 7.6 delivered via PVS (using RAM write-cache).  We desire 32GB RAM per VM.  Physical RAM can be increased.  The VM sizing is the result of testing and recommendations from Citrix.

So as it stands the NUMA nodes are:

1x 10-core Xeon (20 logical procs)
80GB RAM

We want the VMs to be:

8x vCPU
32GB RAM (including 8GB non-paged-pool PVS write cache)

Clearly these servers weren't bought for this specific purpose - 8 doesn't go into 10 (or 20) very well, and 32 doesn't go into 80 very well either! But we have what we have.

So lets say i spin up 4 VMs on this host.  Presumably they will utilise 32 of the 40 logical procs (8vCPU x 4VMs) and 128GB (32GB x 4VMs) of the 160GB of RAM.  I'm guessing ESXi will spread these across both NUMA nodes, so there will be 2x VMs on each socket, which leaves 4 logical processors and 16GB pRAM 'unused' on each of the two NUMA nodes.

If i spin up a 5th VM of the same spec, there is enough physical resource to satsify it however both CPU and RAM will be split equally between the two NUMA nodes right?  4xCPU and 16GB from each to make up the 8xvCPU and 32GB VM.

Presumably i am going to see some performance degredation on this VM that spans?

Is this still measurable and noticeable on a modern host such as this?

NUMA is about memory right?  So if i make sure each socket has enough RAM to satisfy 3 (instead of 2) VMs worth of workload on it, then potentially I could have 3 VMs localised on each NUMA node, with a modicum of over-comittment on the CPU (24vCPUs running on 20 logical CPUs)?  Our Citrix workload is far more RAM-heavy than CPU-heavy.

Any thoughts or advice?

Thanks.

Dave