I also have an M1 with 16GB, what are you doing that you feel it's "normally under memory pressure"? I don't feel any memory pressure but I might not be able to see the signs, or have a different workflow (if you regularly edit photos, videos, 3D, ML, etc. then _of course_ the more RAM the better, but for normal webdev?)
If you open the Activity Monitor and switch to the memory tab you'll see a graph of the memory pressure at the bottom. I think it roughly reflects the degree to which the system is swapping out to the SSD?
On M2 16GB I recently accidentally ended up with Firefox, Chrome, Thunderbird, Spotify, Intellij, GoLand, VSCode, and Docker doing heavy builds in a cross-platform VM. Obviously unreasonable, but interestingly the system crashed with an SSD error.
I only use Docker for a compilation environment for Yocto, luckily I was able to find someone who had instructions for recompiling the image to aarch64 which runs super fast and efficiently.
It was horrible trying to run emulated, but I was sorting this out using the older version before Rosetta 2 was an option - does that make any difference?
Memory pressure on Mac means you get this popup saying "you must close one of these apps NOW or I'm going to crash". On 8GB that's indeed pretty common.
M-series 8gb MacBook Airs and 8gb MacBook Pros have now been in the market for years. The M1 Air launched in 2020.
Yet you'll still have the usual zealots on HN claiming that they have this exact device and it's horrible and worthless and swear blue that their experience is representative.
If the 8gb Air and Pro were a problem: you'd have heard it loud and clear. It would be on the news and called MemoryGate. TheVerge would dedicate a week on it, then bring it up every time apple launch a new laptop. Android9to5 would ring the figurative Apple deathknell and a flurry of lawsuits would be aiming for class action status.
This entire thread is a laughable nothing burger and it's only served to bring out the liars and those on an anti-apple crusade. Some people here have a brain disease that flares up the moment the word "Apple" is muttered.
No one is saying that the mere existence of the 8GB Macbook is a "problem". But it is disingenuous to claim that 8 GB unified memory = 16 GB regular RAM.
The reason why I write these responses is that I value merit-based discussions. Far too much cheerleading goes on in HN. I've said it numerously: It's very difficult to hold Apple, Google, Microsoft, etc. accountable for actual transgressions when naysayers are just boys that cry wolf. These lot are self-defeating, no one worthwhile can take their opinions seriously because they don't apply them consistently and clearly have an agenda.
>No one is saying that the mere existence of the 8GB Macbook is a "problem".
This very thread is saying it ad nauseam - their issues being that (i) an 8gb computer shouldn't be labelled "pro", and (ii) that the 8gb models are irrelevant for serious work.
But onto the meat of the discussion:
While 8 is clearly not 16, Apple are obviously aware that the two numbers are not equal. Instead, what we have here in HN today is a disturbing number of disingenuous commenters who want to take that most absurd position on the statement.
The claim is evidently about how the user would perceive the performance and memory usage of the newer 8gb M-series macs versus apple's earlier 16gb intel models. It's not an unfair comparison and the article made numerous direct quotes about how apple make this claim; those points are being discussed by exactly zero people in this thread. It's also trivial to prove such a claim, so I doubt the majority of the commenters submitting their thoughts even clicked the article.
On a related topic:
A common speed test for smartphones is to simply run many 3rd party apps and multitask between them. These tests aim to demonstrate the real world performance of the device. Now despite having less ram that comparable flagships iPhones perform fluidly and outpace their Android counterparts, often significantly so.
What this underlines is that the performance of a device is not a mere sum of its basic parts. One can't look at a ram figure across different architectures and have an idea about the memory management of the device, or even its performance. A generation of computer users have grown up in a PC-monopoly and fail to grasp that different architectures lend to different efficiencies. Apple's approach here is enough to revisit that thinking.
> This very thread is saying it ad nauseam - their issues being that (i) an 8gb computer shouldn't be labelled "pro", and (ii) that the 8gb models are irrelevant for serious work.
First of all, there are exactly zero comments under the root comment of this thread that discuss the "pro" naming convention. There are comments discussing this in an adjacent thread but they account for only 10% of the total comments on this post.
Nor are there many comments regarding "8gb models are irrelevant for serious work". They exist, but make up only a small percentage of the comments on this post.
HN is really not as harsh on Apple as you think.
> While 8 is clearly not 16, Apple are obviously aware that the two numbers are not equal. Instead, what we have here in HN today is a disturbing number of disingenuous commenters who want to take that most absurd position on the statement.
Honestly, these people frustrate me too, but luckily as mentioned previously, they are really not all that common.
> The claim is evidently about how the user would perceive the performance and memory usage of the newer 8gb M-series macs versus apple's earlier 16gb intel models. It's not an unfair comparison and the article made numerous direct quotes about how apple make this claim; those points are being discussed by exactly zero people in this thread. It's also trivial to prove such a claim, so I doubt the majority of the commenters submitting their thoughts even clicked the article.
Nobody is discussing this because as you mentioned, the claim is trivial to prove. There is nothing to discuss, we all know the M1/2/3 is amazing!
> A common speed test for smartphones is to simply run many 3rd party apps and multitask between them. These tests aim to demonstrate the real world performance of the device. Now despite having less ram that comparable flagships iPhones perform fluidly and outpace their Android counterparts, often significantly so.
>
> What this underlines is that the performance of a device is not a mere sum of its basic parts. One can't look at a ram figure across different architectures and have an idea about the memory management of the device, or even its performance. A generation of computer users have grown up in a PC-monopoly and fail to grasp that different architectures lend to different efficiencies.
This is all very true, but it still doesn't mean 8GB == 16GB. Look, if you want to say "my device performs better than the other device even though it has less RAM", just say that, because it's true!
Be careful when looking at memory pressure and how much it is using. For the same kinds of things I do, it scales well between 128GB of ram to 32 and even 16 GB of ram. If you are using a 32 GB machine and seeing it has high memory pressure or used up most of the memory, it does not necessarily mean it won’t work well on an 8 GB machine.
To know if 8GB is enough, you often need to just try it and see. So that’s why I don’t want to put myself in that situation (that I’m running out of memory and the only option is to upgrade an entire machine.)
(It is because there’s many things aggressively using more memory to make things faster or more responsive by cache things. But those can be released under high memory pressure without much apparent performance loss. That makes use the memory more opportunistically but making it hard to gauge how much really is needed.)
"Enough" certainly. My newest Macbook Pro is 2013, so I have zero insight into the performance of the Apple Silicon devices, but the nature of software development is memory intensive in my experience. The biggest reason claiming "8GB == 16GB" is silly is that Apple doesn't have control over many of the programs that people are using (much as they would like to).
Maybe Xcode makes spectacular use of their own hardware, but do we believe that these optimizations are present in Chrome and Photoshop and Docker and Emacs (hahaha) and whatever NodeJS tooling and probably more that many developers are using at all times? I really, really doubt it. Unless the memory compression discussed here is capable of 50% reduction on average, then it's just a dumb thing to say.
Ultimately your programs all want to have some readily-available bytes in RAM, and most of them don't just cycle out constantly. 8GB is a hard upper bounds on multitasking, and while it might not be our grandpa's 8GB, I really have a hard time believing that it's comparable to 16GB. All this is is beside the point anyway: I switched to desktop like 5 years ago so I could affordably have 64GB and I could never go back. Turns out I don't actually want to take my work home with me, too.