NUMA and Cluster-On-Die

NUMA implementation has gone through several phases of Development in vSphere. At first, it was only responsible for initial placement of VMs, then its functionality was extended with dynamic balancing. In vSphere 5 VMware has presented support of Wide-VMs by exposing NUMA architecture to virtual machines.

New CPUs have presented additional feature - Cluster-on-Die - of splitting physical CPU sockets into  NUMA domains. 

The Full article can be read here

NUMA - Non Unified Memory Access

Last two days I spent quite a lot of time on reading absolutely new topics for me - NUMA, Transparent Page Sharing, TLB, Large Pages, Page Table, and I hope within this week I will make several posts on each of these technologies and how they work together. I have done some tests playing around with Large Pages in ESXi and would like to share this information as well.

I want to start with an explanation about what the NUMA is and to what extent ESX is NUMA-aware. There is a huge amount of information about NUMA in the Internet and probably you are already aware about it, so please let me know if I made any mistakes in my topic.

NUMA stands for Non Unified Memory Access and has nothing to do with Romanian music band and its song "Numa, Numa, yeah". Currently it is presented in Intel Nehalem and AMD Opteron processors. I always assumed that CPUs equally share memory, however this is not the case with NUMA.