Hacker News Reader

Comment by acr25 | Hacker News Reader

This question seems to ignore how RAM works in the Apple Silicon machines. It's part of a system--as available, as quickly, to the CPU as to the GPU.

View on HN

It does not. OP asks if the increase in speed makes up for the limited quantity. They are asking for people's real world experience with the tradeoff.

yeah, that's a bunch of crap. If you have something that needs lots of RAM (say audio uses where samples and such must be in RAM for latency reasons) swapping to SSD is NOT a solution.

And swapping to SSD is pretty stupid - SSDs have finite write lifetimes. Yeah, you can get data center capable SSDs that are over provisioned and cached to the hilt - but at that point RAM would just be cheaper; and still be faster.

I get why Apple is sticking with the tightly coupled RAM on SOC - it's part of the equation that gets them their blistering performance (love my maxed out M1 Max - but kind of nuts 64GB is the most it can have). Some use cases just need more memory - period.

I'm still hoping they will have a true desktop chip at some time where RAM could be externalized even if it causes a slight performance hit. The problem is the number of users who truly need that much RAM is so small I doubt they can justify diverging from their existing architecture that severely.

  > I'm still hoping they will have a true desktop chip at some time where RAM could be externalized even if it causes a slight performance hit.

I believe it was John Ternus on the WWDC23 episode of The Talk Show that said this isn’t going to happen.

They were uncharacteristically firm about this. Same with supporting external GPUs. The indication was this is not going to happen.

How is that relevant?

Spends a lot fewer cpu cycles waiting for data. Same is true with how much faster the main storage moves. Won’t be enough for all workflows (hence it’s worth the OPs question) but will go further than expected.

Edit-latest i9s seem to get around 90GB/s to ram while they promise the M2 ultra 800GB/s. So quarter second to access all of the ram?

Edit2-the ssd is higher end nvme speeds.

Bandwidth doesn't tell us about how many clock cycles it's waiting for the data though (latency), does it?

For example with paging, I don't think those bandwidths make much difference and it's instead the latency which matters. Or am I wrong?

What’s going to have the faster latency? Memory chiplets on the same substrate as the cpu cores or memory dimms sitting in a housing a few inches away?

I’m sure there’s more than a few factors making latency faster for those chiplets.

https://www.anandtech.com/show/17431/apple-announces-m2-soc-...

Edit-but I don’t think this question is about latency exactly. If you’re editing a large file that fits in 196gb ram all at once then those edits should benefit by apparently an order of magnitude higher bandwidth. It’s going to be able to address all of its ram several times per second while the i9 will take much longer (roughly). That and similar speedups from the whole systems engineering should make up some for not having more ram. Think how 8 gig of ram in the m1 was widely reported to be more usable than expected when announced.

Memory chiplets on the same substrate as the cpu cores or memory dimms sitting in a housing a few inches away?

Where are you getting this information on the memory latency? What is the actual number?

I don’t think this question is about latency exactly.

You were the one the brought up latency

those edits should benefit by apparently an order of magnitude higher bandwidth

Most software is not written well enough to be memory bandwidth limited, even graphics software. It is mostly games that end up actually needing the memory bandwidth.

You mean like how 8GB RAM in my Mini results in prolonged swap pauses with 3 apps open?

To be fair, M1 MacBook Pro with 32GB is smooth as silk with ~20 apps open.

Which 3 apps on the Mini? I’ve actually been really impressed with my base model M1 Mini. I’ve been able to use XCode, Safari, iMessage, and Music playing without many noticeable hiccups at all.

Chrome, PHPstorm, OpenSCAD, Cura. I lied, 4 apps.

My point though is, doesn’t matter how fast your RAM is if you run out and have to swap out.

Oh yeah JetBrains IDEs definitely need more RAM than Xcode. Likewise with Chrome versus Safari. I’m not familiar with OpenSCAD and Cura.

OpenSCAD is a problem even with 96GB ram.

Something is very wrong with their build on Apple Silicon. It works fine for me on other platforms.

I've been having acceptable experience with it on my 32GB M1 MBP with the Apple Silicon build. I admit though, I don't use it as much on my work laptop as my home mini.

Reply on news.ycombinator.com