Talk:NVDIMM

info
http://www.snia.org/sites/default/files/NVDIMM%20Enabling%20Greater%20ROI%20from%20SSDs.pdf — Preceding unsigned comment added by FxJ (talk • contribs) 22:40, 22 June 2016 (UTC)

NVSRAM VS NVDIMM Power consumption
I am looking for information NVSRAM VS NVDIMM Power Consumption and cannot find it. — Preceding unsigned comment added by 193.108.195.249 (talk) 06:45, 28 June 2016 (UTC)

Intro rewritten
I rewrote the first paragraph of the Intro, tightening up the wording, providing a better introduction, stating benefits and drawbacks earlier, and eliminating phrasing such as "Some consider...misleading." The remainder of the Intro, introducing NVDIMM-F and NVDIMM-N, does not make a clear enough distinction, as both types seem to pair volatile and non-volatile memories. (In fact, if they all do, then my use of the word "may" should be corrected.) Spike-from-NH (talk) 15:58, 17 January 2017 (UTC)

Rebuttal
"Non-volatile" products may use volatile memory during normal operation and dump the contents into non-volatile memory if the power fails, using an on-board backup power source. Volatile memory is faster than non-volatile; it is byte-addressable; and it can be written to arbitrarily, without concerns about wear and device lifespan. However, including a second memory to achieve non-volatility (and the on-board backup power source) increases the product cost compared to volatile memory.

The above is not really the informational standard of a highly technological lead, and it is also somewhat misleading, because volatile technology does not have a secure stranglehold on that feature quadrant. There's a vocal and active claimant to the throne presently claiming all the same things.

My rebuttal (with citation):

Of the many non-volatile memories in the development pipeline, Nano-RAM based on carbon nanotubes is purported to come closest to DRAM on the criteria of performance, byte-addressability and device lifespan; first products are expected in 2021 at moderate density, from fabrication partner Fujitsu. If this ambitious technology pans out, and then manages to scale cost-effectively in the near horizon, it's not clear that DRAM has no direct non-volatile replacement.

But really, the correct fix is to drain some attitude out of the preceding paragraph and expand the lead out with more concrete, factual material. &mdash; MaxEnt 16:20, 19 September 2020 (UTC)