Server requirements for running model for Churn Management for 150,000 mobile customers - Will 64 Core, 512 GB RAM and 4 TB storage enough? Number of servers / VM required? Also how will VM be defined

yes

it depends upon the Server capacity and provisioning has to be defined

I have ran a churn model for 15million subs on a normal laptop before and it didn’t give any headache. So with this smaller base you can go with the storage specs you have indicated.

Should be more than sufficient.
However you didn’t say the time series for the churn model so 150M over 10 years, i’d segment the data to joiners and leavers. The model also needs to include both identical identity matches(e.g. same name/address, same name/credit card details, same address/credit card details) as well as similar identities.(e.g. same name same locale).
Note that in preprocessing identity that cleanse, translate and standardisation are required on name and address.

The system seems powerful enough for models .

Yes, it will do. And I must say with 512 GB RAM you can go almost anywhere in the world for now…

Should be more than sufficient.

Yes! That will be more than enough.

it also depends on the frequency of refresh of model output. Will the model be running in real time for over a day or week. The processing is what requires resoruces

Just use Google Colab it’s free.. you can get a machine for free with GPU and 12gb mem.. More than you need for 150K customers..

That seems like crazy over-spec - best bet would be to experiment, how often do you want to run the models and how fast do you need results, I can’t see why you’d want this running 24/7 - might be something you’d want to bring up run and shut down in a VM or possibly even containerized

To run a churn management model for 150,000 mobile customers, the server requirements will vary based on several factors such as model complexity, data set size, and usage patterns. However, in general, a server with 64 cores, 512 GB RAM, and 4 TB storage can handle the workload efficiently, provided the model is well-designed.

To determine the number of servers/VMs required, several factors like expected traffic, fault tolerance level, and desired performance metrics need to be considered. Multiple servers/VMs are recommended for workload distribution and redundancy in case of failure. Using two or three servers/VMs initially and monitoring their performance and resource utilization can be a good starting point.

Defining the VM depends on the virtualization platform utilized. A VM comprises virtualized CPU, RAM, storage, and network interface. Virtual CPUs, RAM, and virtual disk space should be set based on the workload and expected usage patterns. The virtual network interface should be configured to offer adequate bandwidth and network connectivity. It is advisable to consult the documentation of the virtualization platform to identify the optimal configuration for the VM.

Good day, the specified VM will be sufficient for churn management and it should be auto scalable in the event of increased capacity is required although it depends on the budget requirements for this solution.

Virtual machine is a software driven infrastructure. Within the category of infrastructure as a solution (IaaS) that are available in d cloud computing environment.  Where you can request any specifications and capacity in terms of hardware and software resources anyone may need