Editor !
2u latency .. what ??
Its 2us latency ... you need to specify the units that are being micro'ed.
Early-stage startup Xitore says it is developing bigger and faster non-volatile flash DIMMs to give servers a performance boost so they could run more VMs and containers and run them faster by sidestepping high latency storage IOS. It is taking up the baton brought onto the field by Netlist and Diablo Technologies and …
The slide says 2us.
While this sounds good... 4TB per DIMM, the problem is that its still vaporware.
If it works... 4TB per dimm. Consider a Supermicro motherboard that has 24 Dimm slots.
Its possible to put 1TB of DDR4 Memory and then put in 40TB in the extra slots.
(YMMV based on memory density. )
Or if you want to scale down... a blade w one or two == 8TB of storage per card in a blade server.
No need for SSDs or a small M.2 Flash drive for OS and Swap.
If real, this could be a game changer in the short term.
Sorry if I seem skeptical, but we've seen a lot of hype in this space.