Have any examples from the past decade? Especially in the context of how exaggerated the claims are from PC and Android brands they are competing with?
Fair point. Based on their first sentence, I mischaracterized how “laughable” was used.
Though the author also made clear in their second sentence that they think Apple is one of the worst when it comes to marketing claims, so I don’t think your characterization is totally accurate either.
Ye that was hilarious, my basic workload borders on the 8GB limit not even pushing it.
They have fast swap but nothing beats real ram in the end, and considering their storage pricing is as stupid as their RAM pricing it really makes no difference.
If you go for the base model, you are in for a bad time, 256GB with heavy swap and no dedicated GPU memory (making the 8GB even worse) is just plain stupid.
This what the Apple fanboys don't seem to get, their base model at somewhat affordable price are deeply incompetent and if you start to load it up the pricing just do not make a lot of sense...
> If you go for the base model, you are in for a bad time, 256GB with heavy swap and no dedicated GPU memory (making the 8GB even worse) is just plain stupid ... their base model at somewhat affordable price are deeply incompetent
I got the base model M1 Air a couple of years back and whilst I don't do much gaming I do do C#, Python, Go, Rails, local Postgres, and more. I also have a (new last year) Lenovo 13th gen i7 with 16GB RAM running Windows 11 and the performance with the same load is night and day - the M1 walks all over it whilst easily lasting 10hrs+.
Note that I'm not a fanboy; I run both by choice. Also both iPhone and Android.
The Windows laptop often gets sluggish and hot. The M1 never slows down and stays cold. There's just no comparison (though the Air keyboard remains poor).
I don't much care about the technical details, and I know 8GB isn't a lot. I care about the experience and the underspecced Mac wins.
I don't know about your Lenovo and how your particular workload is handled by Windows.
And I agree that in pure performance, the Apple Silicon Macs will kill it; however, I am really skeptical that an 8GB model would give you a better experience overall.
Faster for long compute operations sure, but then you have to deal with all the small slowdown from constant swapping. Unless you stick to a very small amounts of apps and very small amounts of tabs at the same time (which is rather limiting) I don't know how you do it.
I don't want to call you a liar but maybe you are emotionally attached (just like I am sometimes) to the device to realize it, or maybe the various advantages of the Mac make you ignore the serious limitations that come with it.
Everyone has their own sets of tradeoffs but my argument is that you can deal with the 8GB Apple Silicon devices you are very likely to be well served by a much cheaper device anyway (like half as cheap).
All I can say is I have both and I use both most days. In addition to work-issued Windows laptops, so I have a reasonable and very regular comparison. And the comparative experience is exactly as I described. Always. Every time.
> you have to deal with all the small slowdown from constant swapping
That just doesn't happen. As I responded to another post, though, I don't do Docker or LLMs on the M1 otherwise you'd probably be right.
> Unless you stick to a very small amounts of apps and very small amounts of tabs at the same time
It's really common to have approaching 50+ tabs open at once. And using Word is often accompanied by VS Code, Excel, Affinity Designer, DotNet, Python, and others due to the nature of what I'm doing. No slowdown.
> maybe you are emotionally attached
I am emotionally attached to the device. Though as a long-time Mac, Windows, and Linux user I'm neither blinkered nor tribal - the attachment is driven by the experience and not the other way around.
> maybe the various advantages of the Mac make you ignore the serious limitations that come with it
There are indeed limitations. 8GB is too small. The fact that for what I do it has no impact doesn't mean I don't see that.
> you can deal with the 8GB Apple Silicon devices you are very likely to be well served by a much cheaper device anyway (like half as cheap)
I already have better Windows laptops than that, and I know that going for a Windows laptop that's half as cheap as the entry level Air would be nothing like as nice because the more expensive ones already aren't (the Lenovo was dearer than the Air).
---
To conclude, you have to use the right tool for the job. If the nature of the task intrinsically needs lots of RAM then 8GB is not good enough. But when it is enough it runs rings around equivalent (and often 'better') Windows machines.
Not individually, no. Though it's often done simultaneously.
That said you're right about lots of RAM in that I wouldn't bother using the 8GB M1 Air for Docker or running LLMs (it can run SD for images though, but very slowly). Partly that's why I have the Lenovo. You need to pick the right machine for the job at hand.
You know that RAM in these machines is more different than the same as "RAM" in a standard PC? Apple's SoC RAM is more or less part of the CPU/GPU and is super fast. And for obvious reasons cannot be added to.
Anyway, I manage a few M1 and M3 machines with 256/8 configs and they all run just as fast as 16 and 32 machines EXCEPT for workloads that need more than 8GB for a process (virtualization) or workloads that need lots of video memory (Lightroom can KILL an 8GB machine that isn't doing anything else...)
The 8GB is stupid discussion isn't "wrong" in the general case, but it is wrong for maybe 80% of users.
> EXCEPT for workloads that need more than 8GB for a process
Isn't that exactly the upthread contention: Apple's magic compressed swap management is still swap management that replaces O(1) fast(-ish) DRAM access with thousands+ cycle page decompression operations. It may be faster than storage, but it's still extremely slow relative to a DRAM fetch. And once your working set gets beyond your available RAM you start thrashing just like VAXen did on 4BSD.
Exactly!
Load a 4GB file and welcome the beach ball spinner any time you need to context switch to another app.
I don't know how they don't realize that because it's not really hard to get there.
But when I was enamored with Apple stuff in my formative years, I would gladly ignore that or brush it off so I can see where they come from, I guess.
It's not as different as the marketing would like you to think. In fact, for the low-end models even the bandwidth/speed isn't as big of a deal as they make it out to be, especially considering that bandwidth has to be shared for the GPU needs.
And if you go up in specs the bandwidth of Apple silicon has to be compared to the bandwidth of a combo with dedicated GPU. The bandwidth of dedicated GPUs is very high and usually higher than what Apple Silicon gives you if you consider the RAM bandwidth for the CPU.
It's a bit more complicated but that's marketing for you. When it comes to speed Apple RAM isn't faster than what can be found in high-end laptops (or desktops for that matter).
Not sure what windows does but the popular method on e.g. fedora is to split memory into main and swap and then compress swap. It could be more efficient the way Apple does it by not having to partition main memory.
I do admit the "reliance on swap" thing is speculation on my part :)
My experience is that I can still tell when the OS is unhappy when I demand more RAM than it can give. MacOS is still relatively responsive around this range, which I just attributed to super fast swapping. (I'd assume memory compression too, but I usually run into this trouble when working with large amounts of poorly-compressible data.)
In either case, I know it's frustrating when someone is confidently wrong but you can't properly correct them, so you have my apologies
I suggest you go and look HOW it is done in apple silicon macs, and then think long and hard why this might make a huge difference. Maybe Asahi Linux guys can explain it to you ;)
I understand that it can make a difference to performance (which is already baked into the benchmarks we look at), I don't see how it can make a difference to compression ratios, if anything in similar implementations (ex: console APUs) it tends to lead to worse compression ratios.
If there's any publicly available data to the contrary I'd love to read it. Anecdotally I haven't seen a significant difference between zswap on Linux and macOS memory compression in terms of compression ratios, and on the workloads I've tested zswap tends to be faster than no memory compression on x86 for many core machines.
