amd ryzen 9 7950x3d how many vcpu can it make for shared vcpu vms like vultr

amd ryzen 9 7950x3d how many vcpu can it make for shared vcpu vms in KVM like vultr
i m doing some research to provide top-notch vms on the latest hardware so please help me understand
the hardware i will use is
CPU: AMD Ryzen 9 7950X3D (5.7 GHz)
RAM: 128 GB DDR5 ECC 4200 mhz
Disk: 9.60 TB NVMe gen 4 nvme pcie
so i need some help to know haw many vcpu i can make to provide vcpe better than vultr cuz this vcpu is Base Clock 4.2GHz boosted to 5.7Ghz
and wich location is best i have to choose 3 location for now i think of
poland in europe and london in uk and dallas or newyork in usa is that good and please any recomendation like what you need and thanx

Tagged:

Comments

  • i still didnt understand how many shared vcpu like vultr they provide dedicated thread as vcpu from plans over 30usd but in their 6$ plans like does they split one thread to 3 vcpu with fair share or how many

  • edited June 2023

    @gaithads said:

    they split one thread to 3 vcpu with fair share or how many

    Depending on use case

    Your can experiment and see what the usage is going to be.

    Top notch is subjective.
    Does it mean there is cpu steal?
    Are your customers going to be OK with slower than advertised cpu?

  • is the performance of general purpose 5 usd vcpu ios it ok to get 3vcpu per one dedicated thread?

  • It depends what's the usage of customer vm.

    You want to strike a good balance and monitoring to ensure threads do not get abused if shared.

    I feel you are trying to get the optimal number first before diving in, but sometimes it's OK to start with a guesstimate and fine tune it in future.

    Let customers know ahead of time that the cpu policies may change and are not written in stone.

    If you are offering shared cores make sure you are able to find out abusers (there will always be one and more) and take action

    Plus you can start conservative by keeping some threads free and offer only a few to users and check how that works

    Good luck

  • the server is cost a lot cus its latest tech and high frequency and ram etc thats why i want to seeif the total vms can payoff the server and make some revenu thats why and i want to give good experience

  • @gaithads said:
    the server is cost a lot cus its latest tech and high frequency and ram etc thats why i want to seeif the total vms can payoff the server and make some revenu thats why and i want to give good experience

    You need to do some math to find that out. No one can tell you a definitive answer to the question because of all the factors involved.

    Only you know those factors we never possibly could even guess because it is your setup not ours.

    We could debate theories on this subject all day and night and you still will not an answer.

    Thanked by (2)FrankZ tetech

    URL Shortener | YetiNode | Come join us on the MetalVPS IRC channel!!! | Don't be a jerk, let YetiNode do the work.

  • edited June 2023

    The bottom neck likely be the memory than CPU itself. Remote server softwares that are high CPU intensive tend to be high on memory usage, you likely run out of memory first before running out of CPU.

    Thanked by (1)tetech
  • @lu5ck said: The bottom neck

    You mean like a sphincter?

Sign In or Register to comment.