I went with 6x 64GB LRDIMMs for 384GB total. I'm happy I went with six. MEM-DR464L-HL03-LR26. *edit: corrected part number to HL03
Gotta have a photos after spending $:

the labels on the plastic were smudged because I sprayed them with lysol. LOL.

I now have two used 32GB 2666MHz DDR4 RDIMMS back in a box waiting to be used again some future day:

about 9 months ago I enabled Transparent Page Sharing (TPS) for deduplicating memory pages and also disabled large memory pages (smaller memory pages increased chances of being able to deduplicate those smaller pages between VMs). This article explains it well: VMware memory optimization
mem.ShareForceSalting = 0 *enables TPS on all VMs of the host
Mem.AllocGuestLargePage = 0 * disables large memory pages
The above settings would keep my consumed host memory below 64GB while running my necessary guest VMs. Without TPS the VMs that were running were consuming too much RAM and balooning would occur followed by swap files and dog slow performance.
But now that I have plenty of RAM at the moment (for now) I've reset those values to the following:
mem.ShareForceSalting = 2 *disables TPS on all VMs of the host
Mem.AllocGuestLargePage = 1 * enables large memory pages (advantage is better performance relating to MMU but at cost more RAM being consumed by each VM because larger pages means less chance of deduplication if using TPS)
If you have enough RAM do you still enable TPS (VM page sharing) and small memory pages?... or do you prefer large memory pages and therefore no page sharing but having greater MMU performance?
Thanks
Gotta have a photos after spending $:

the labels on the plastic were smudged because I sprayed them with lysol. LOL.

I now have two used 32GB 2666MHz DDR4 RDIMMS back in a box waiting to be used again some future day:

about 9 months ago I enabled Transparent Page Sharing (TPS) for deduplicating memory pages and also disabled large memory pages (smaller memory pages increased chances of being able to deduplicate those smaller pages between VMs). This article explains it well: VMware memory optimization
mem.ShareForceSalting = 0 *enables TPS on all VMs of the host
Mem.AllocGuestLargePage = 0 * disables large memory pages
The above settings would keep my consumed host memory below 64GB while running my necessary guest VMs. Without TPS the VMs that were running were consuming too much RAM and balooning would occur followed by swap files and dog slow performance.
But now that I have plenty of RAM at the moment (for now) I've reset those values to the following:
mem.ShareForceSalting = 2 *disables TPS on all VMs of the host
Mem.AllocGuestLargePage = 1 * enables large memory pages (advantage is better performance relating to MMU but at cost more RAM being consumed by each VM because larger pages means less chance of deduplication if using TPS)
If you have enough RAM do you still enable TPS (VM page sharing) and small memory pages?... or do you prefer large memory pages and therefore no page sharing but having greater MMU performance?
Thanks
Last edited: