[boojums]

DRAM/NAND processes are optimized for different things than logic, so there isn't much cross application. My understanding is that it would be hard to make a general compute chip with a DRAM process due to a low number of metal layers, transistor types, etc. Micron [1] and Samsung have both investigated doing massively parallel compute in the memory cell, but the technology never panned out.

Regarding EUV, according to Micron's most recent quarterly report (March 20th) [2]

"We continue to mature our production capability with extreme ultraviolet lithography (EUV), and have achieved equivalent yield and quality on our 1α as well as 1ß nodes between EUV and non-EUV flows. We have begun 1γ (1-gamma) DRAM pilot production using EUV and are on track for volume production in calendar 2025."

[1] https://investors.micron.com/news-releases/news-release-deta...

[2] https://investors.micron.com/static-files/1a8d6c22-3b89-4806...

[samcheng]

What about Apple's M-series chips, which have on-die RAM? Is that RAM significantly more expensive per GB due to the more expensive process?

(It's certainly exorbitantly expensive for retail consumers at $200 for an 8 GB RAM upgrade on a Macbook!)

[wmf]

Apple does not have on-die RAM. The SoC is a normal logic die and the RAM is regular DRAM.

[samcheng]

Ah - I see, the DRAM is literally BGA'd right next to the die. So that $200 upcharge is mostly profit... Thank you!

[vbezhenar]

It's called "package on package". The RAM is different chip, however it's located very close to the CPU chip and both are under a single cover. The end result is a "package".

I think that GPUs use similar approach.

[RF_Savage]

AMD Vega used HBM next to the GPU die, but most GPU manufacturers are now back to discrete chips on the GPU circuit board.