Gregory's ServerFerroelectric RAM (FRAM) is an unusual type of memory. It is fast and can store data for decades without power. I opened up a FRAM chip to reveal the tiny cubes of PZT, the ferroelectric material that holds the bits.
 23 comments
 This die photo shows the 64-kilobit Ramtron chip (FM24C64). The memory is partitioned into four rectangles, each holding 32768 tiny cubes of PZT. The chip is accessed serially (using I2C), so it only has eight pins; you can see the bond pads around the edges of the die. Here's a closeup of the sense amplifiers. The signals from the capacitors are very small, so each bit is stored in two capacitors, one high and one low. This makes it easier to distinguish a 0 and a 1. The sense amplifier boosts these two signals to determine a 0 or a 1. (DRAMs use similar sense amplifiers to read bits.) Because reading an FRAM capacitor destroys its value, the sense amplifier's output is then written back to memory. To make a capacitor, each PZT cube has a platinum plate line underneath and a platinum plate contact on top. The plate lines are the shiny vertical white rectangles in this photo. At the bottom, large transistors drive the selected plate line positive or negative. To learn more about this FRAM chip, see my blog post: https://www.righto.com/2024/09/ramtron-ferroelectric-fram-die.html  @bitsavers There are multiple factors that limit the number of writes that FRAMs can handle: changes in crystal structure as Ti ions replace O, mobile ions collecting at grain boundaries, and something to do with 90º domains.  @kenshirriff @bitsavers it's not really infinite since the endurance was probably reached in Marc's DRO. The fram was used by the readout to give the illusion of an absolute scale, but I believe the integrated fram that failed was updated too often (maybe every scale tick).   @kenshirriff When you read the capacitor you do discharge it like in a dynamic ram ? Is it a destructive read that needs the value to be restored ? @gilesgoat Yes, reading the capacitor is destructive. You force the capacitor to the 1 state; it takes more current to do this if it was in a 0 state initially. (It is very much like magnetic core memory, with a hysteresis curve.)  @kenshirriff It's interesting that it's a modern tech. It feels like one of those concepts that would have been commercialized in the 60s and now be out of fashion. Abbreviating lead as P seems a bit like taking the biss. @wikkit Ferroelectric RAM dates back to the early 1950s, strangely enough. This photo from Scientific American, 1955, shows a 256-bit memory constructed by Bell Labs. Abbreviating lead as P kind of makes sense since lead is Pb; all three elements lose their second letter in PZT. Oxygen got dropped entirely from the abbreviation. @kenshirriff I read the wikipedia article about the inventor, he seemed super bright and I have to wonder what he would have done with the second half of his life.  @kenshirriff TFW I realize these are actually commercially available for a few bucks per 64kb.   @kenshirriff It's good stuff! I use them to eliminate the need for the 3 AA batteries seen in so many pinball machines. With the FeRAM installed, I can just remove the battery holder and I've never had any issues.  @kenshirriff Are these still used in the present day and, if so, what's the use case? I had no idea FRAM existed until now.  @kenshirriff This brings back memories of an IBM 360 derivative machine from the 1970s that had memory modules made of tiny magnetic core donuts with tiny wires orthogonal to each other. Maybe they were 8K each but the improved ones were 16K I think.  But with those capacities, we're back to having full-length RAM adapter cards in our systems, eh? I miss those days! |
The memory uses capacitors built from PZT ( lead zirconate titanate), a ferroelectric material. The zirconium or titanium atoms inside each crystal cell move up or down, causing the capacitor to get "stuck" in the positive or negative state, holding a 1 or a 0.