Yet another Hacker News

New CRAM method offers giant boost to compressed memory reads

39 points danny00 7 hours ago 18 comments

https://lpc.events/event/20/contributions/2424/

https://www.phoronix.com/news/Linux-CRAM-Compressed-RAM

Kevin_Flynn 7 hours ago | parent

Ram Doubler, for Mac and Win 3.x, 1996

https://winworldpc.com/product/connectix-ram-double/windows-...

Nice to see its back, with better performance.

History repeated, with refinement.

vfosnar 5 hours ago | parent

https://www.phoronix.com/news/Linux-CRAM-Compressed-RAM

x0z 4 hours ago | parent

Thank you <3

dang 1 hour ago | parent

Added above. Thanks!

eglintondust 5 hours ago | parent

So can I finally download more RAM?

smallmancontrov 1 hour ago | parent

https://downloadmoreram.com/

Whoa, new options! I'm downloading the Quantum RAM with Haptic Feedback.

swed420 1 hour ago | parent

You wouldn't download a car

paimapi 18 minutes ago | parent

has anybody tried 3D-printing a car yet

skavi 22 minutes ago | parent

probably not to your current computer. this works on supporting memory types that do compression in hardware.

madduci 5 hours ago | parent

Finally we can run frontier models locally

throawayonthe 3 hours ago | parent

one would hope model weights are already high entropy enough and would not be improved by a general-purpose memory compression algo?

sroussey 1 hour ago | parent

We brute force AI models right now because a) we don’t know better, and b) it’s premature optimization.

I beg to differ on point b, but no one is delaying their next model just so they can concentrate on optimization.

It’s coming though.

One example: https://siliconangle.com/2026/07/28/ai-model-compression-sta...

Another is separating the the intelligence part of the model from the known facts part of the model (which can be better compressed)

tintor 1 hour ago | parent

Compression doesn't really work for model weights.

Model quantization and model distillation are two techniques to reduce model size.

cwillu 5 hours ago | parent

The article appears to misunderstand the point of the project, and mostly just describes what zswap already does (writing compressed pages back into ram), rather than talking about cram's hardware-offloaded compression that allows cacheline-level access rather than page-level. (And they appear to be aware that they don't really understand it: “My explanation of CRAM might not be completely correct”)

The phoronix article is better, and I say that as someone who usually detests the quality of phoronix's technical writing.

tancop 2 hours ago | parent

It's not really about hardware offload, the biggest problem with zswap is that it's swap. The rest of the kernel treats it like a fast SSD (which is still incredibly slow compared to RAM) instead of slower memory that needs a bit of special handling on writes.

NUMA maps a lot closer to what compressed RAM actually is. The subsystem is more aware of CRAMs specifics so it can make better decisions about where to put allocations and everything gets faster. And it's less overhead because swap is not really optimized for frequent direct access but for NUMA it's the most basic function.

cyberclimb 3 hours ago | parent

the recording of the presentation has got to be on this YT channel but I only scrubbed through 2 videos (8h each) and it's not easy to find

https://www.youtube.com/@LinuxPlumbersConference

not sure where it falls in the schedule here either https://lpc.events/event/20/timetable/#all

tux3 53 minutes ago | parent

Unfortunately the recording is missing audio for half the talk. A frequent issue for the plumbers livestreams this year.

https://www.youtube.com/live/OPRciCSsdS4?t=2031

cyberclimb 29 minutes ago | parent

Legend! Thanks for digging it up