Comparison of Framework RAM and Regular SODIMM RAM?

Has anyone tested the RAM against any other brands, how well it performs, etc…?

1 Like

Aside from the questionably useful (thus far) official list, someone on Reddit has made a summary (second sheet) of the reports of working third-party RAM modules there and on this forum.

The same as everything else. There should be little to no difference in performance compared to identical kits. Specifically, they should perform very similarly to ADATA kits seeing that it is the memory modules used.

So there is only a list of what is working, not how the memory ranks in terms of performance?

As Water261 mentioned, there really shouldn’t be an appreciable difference between two different RAM kits of similar specs from different manufacturers. Two kits rated at the same speed and latency should perform the same.

Unlike SSDs, which have complex controllers and custom firmware that dictate the performance, RAM is “dumb”. As folks in the thread noted, the controller complexity is all on the processor side, so memory set to the same timings will perform the same. We use JEDEC standard timings, so our DDR5 performs the same as other DDR5-5600 memory using JEDEC timings.

7 Likes

“JEDEC timings” isn’t super precise.

JEDEC offers 4 timing classes for modules. IIRC someone shared that the Framework modules are using JEDEC timing class B.

Timing class B will perform worse than timing class AN (which is available on some kits from G.Skill and Kingston) but better than timing class BN or C.

Another thread where this was being discussed: There are three JEDEC speed bins for DDR5 5600. The highest has 20% lower latency than the lowest. Which is framework selling? (though not answered yet).