@kieran Liquid nitrogen is not needed. Based on https://s3.amazonaws.com/citpsite/wp-content/uploads/2019/01/23195456/halderman.pdf, “canned air” which costs less than $20 will work.
3.2 Decay at reduced temperature
It has long been known that low temperatures can signifi-
cantly increase memory devices’ retention times [29, 2, 46, 23, 41, 40]. To measure this effect, we performed a second series of tests using machines A–D.
In each trial, we loaded a pseudorandom test pattern into memory, and, with the computer running, cooled the memory module to approximately −50◦C. We then powered off the machine and maintained this temperature until power was restored. We achieved these temperatures using commonly available “canned air” duster products (see Section 4.2), which we discharged, with the can inverted, directly onto the chips.
As expected, we observed a significantly lower rate
of decay under these reduced temperatures (see Table 2).
On all of our sample DRAMs, the decay rates were low enough that an attacker who cut power for 60 seconds would recover 99.9% of bits correctly.
As an extreme test of memory cooling, we performed
another experiment using liquid nitrogen as an additional
cooling agent. We first cooled the memory module of Machine A to −50◦C using the “canned air” product. We then cut power to the machine, and quickly removed the DRAM module and placed it in a canister of liquid nitrogen. We kept the memory module submerged in the liquid nitrogen for 60 minutes, then returned it to the machine. We measured only 14,000 bit errors within a 1 MB test region (0.17% decay). This suggests that, even in modern memory modules, data may be recoverable for hours or days with sufficient cooling.