Thanks for your reply and the link to some articles. However, the documents linked and the information provided do not answer the question I have, hence the post.
For context, I have been brought in as a consultant to investigate and resolve performance issues with Qlik View, and I was selected to do this work because I used to be an employee of VMware and so I have a deep and complete knowledge of virtualisation. The current Qlik View servers are a P2V conversion of physical machines, and have been allocated 8 sockets, which is directly recommended against within the "Recommended Top Performing Servers" article, as I linked and quoted in my original post.
Back to the original question, asked in a different way - is there any knowledge or metrics on known performance differences between multi-core versus multi-socket systems? If I switch from an 8 socket (with only one 'core' per socket) to 2 socket (with 4 cores per socket) - on the same physical hardware, will this make a noticeable difference?
