CPU intensive VPS/dedicated servers (80+cores) , low ram/disk/net, or dedicated core(s)

rincewind

BTW, Vultr cores are Haswell. Maybe E3. So you could gain some performance by recompiling with -march=haswell

AVX_512

That does not answer the question at all, "cores" in the same or also different systems?

Yes it scales perfectly between systems. Each workunit is 100% independent with zero cross talk needed between workunits. This project uses a beanstalk queuing server and python script to manage the workunits. I am running many 1 core VPS (Vultr) so my fault for thinking that was clear.

AVX_512

@rincewind said:
BTW, Vultr cores are Haswell. Maybe E3. So you could gain some performance by recompiling with -march=haswell

Thank for that news. I was thinking it had to be at least Ivy Bridge based on the increased performance over the e5-2670 v1 systems but haswell v3 based makes more sense.