Live data from Hacker News

Apple has a memory problem and we're all paying for it

macworld.com

131–140 of 409 posts

Re: Apple has a memory problem and we're all paying for it

#131

Earlier quoted context omitted.

I have a 2011 Thinkpad T520 which still works fine, upgraded it to 16GB of RAM. There are high quality Windows laptops out there, but of course they cost as much as a Macbook. I'm okay with laptops being sold with 8GB of RAM, but I'm not okay with artifically shortened lifespan caused by lack of upgradability. Manufacturers should be regulated to force them offer upgradeable RAM. I don't care what new technological i…

> Manufacturers should be regulated to force them offer upgradeable RAM. Disagreed. You can't make RAM upgradeable and have unified memory architecture that makes Apple Silicon good. Let's not regulate technology like this. There were studies done that showed most people do not upgrade their RAM/storage even though they could. They just bought new computers. I can't find them now but you could probably find them if y…

> You can't make RAM upgradeable and have unified memory architecture

Silicon Graphics did it in the 90s, x86 laptops do something similar but it's not true UMA AFAIU mostly due to software / graphics API limitations.

> that makes Apple Silicon good.

Good interplay between CPU and GPU, the major advantage of UMA, does nothing for raw CPU performance. What I find attractive about Apple silicon is that it has very high performance, high efficiency CPU cores. Nothing to do with UMA.

Soldered RAM can be somewhat higher speed and lower energy than replaceable, I think that's the main real advantage. Still, I wouldn't want it.

Re: Apple has a memory problem and we're all paying for it

#132

Earlier quoted context omitted.

No it's not. A ton of programs used by regular people (Chrome etc.) are RAM hogs, if you are a dev, then Docker is also an egregious example. If you happen to play games (a big thing for average people), that 8GB will be shared between the CPU and the GPU. Even the Steam Deck, admittedly a budget device, comes with 16GB. It's often mentioned that macOS is more frugal with RAM, a claim which I flat out found to be unt…

> No it's not. A ton of programs used by regular people (Chrome etc.) are RAM hogs, if you are a dev, then Docker is also an egregious example. If you happen to play games (a big thing for average people), that 8GB will be shared between the CPU and the GPU. Even the Steam Deck, admittedly a budget device, comes with 16GB. That's why I said normal office workers and students. I never said 8GB is enough for devs. It w…

Thing is, for regular office use, the M1 Air is perfectly fine, there's very little you gain if you go with the M3 Pro instead of the M1.

Funnily enough, I'm typing this from a 3 year old Thinkpad that I use for work, which has in all likelyhood a slower CPU than even the M1, but it's perfectly usable for dev work, because it has 64GB of RAM.

Re: Apple has a memory problem and we're all paying for it

#133
post #114

Earlier quoted context omitted.

Getting by today, doesn't mean one gets by tomorrow.

Why, are we waiting something drastic in the next 4-5 years before the users gets the new model?

Because that is forced obsolence and damaging for the environment, the computer from today should be perfectly usable in 4-5 years.

The upgrade treadmill from the 20 century computers is no longer relevant for most users, besides storage.

Re: Apple has a memory problem and we're all paying for it

#135

Earlier quoted context omitted.

> Manufacturers should be regulated to force them offer upgradeable RAM. Disagreed. You can't make RAM upgradeable and have unified memory architecture that makes Apple Silicon good. Let's not regulate technology like this. There were studies done that showed most people do not upgrade their RAM/storage even though they could. They just bought new computers. I can't find them now but you could probably find them if y…

> You can't make RAM upgradeable and have unified memory architecture Silicon Graphics did it in the 90s, x86 laptops do something similar but it's not true UMA AFAIU mostly due to software / graphics API limitations. > that makes Apple Silicon good. Good interplay between CPU and GPU, the major advantage of UMA, does nothing for raw CPU performance. What I find attractive about Apple silicon is that it has very high…

>x86 laptops do something similar but it's not true UMA AFAIU mostly due to software / graphics API limitations.

They don't. They pre-allocate a portion of the memory to the iGPU and the CPU can't access that memory. It's not nearly the same unified memory we're talking about.

>Good interplay between CPU and GPU, the major advantage of UMA, does nothing for raw CPU performance. What I find attractive about Apple silicon is that it has very high performance, high efficiency CPU cores. Nothing to do with UMA.

I didn't imply that unified memory is the only thing that makes Apple Silicon good. Of course everything else has to be good too. But unified memory is a big part of Apple Silicon. Furthermore, going forward, UMA has real advantages over non-UMA architectures for LLMs.

Re: Apple has a memory problem and we're all paying for it

#136
post #127

Earlier quoted context omitted.

Well, Apple has been using SSDs for a long time now. Do you know Mac users personally where the SSD died before something else in the computer?

Yes I have seen this. The disk write speed randomly gets super slow and tons of stuff just locks up at random as many many programs assume it'll be fine to do little writes synchronously. The whole machine becomes totally unusable. I see it on computers (and phones too) whose owners bought the smallest capacity to avoid spending money on something they didn't think they would use.

That's not what happens when an SSD reaches the write limit. It'll simply not let you write more data and becomes read only.

Re: Apple has a memory problem and we're all paying for it

#137

Earlier quoted context omitted.

What on earth are you doing to your laptops that has you convinced that six years is an exceptionally long lifespan!?!? Anyways even though it's literally not a Windows laptop my system 76 lemur ultra i bought in May 2013 works fine after a decade of regular use with no hinge or screen problems. I came close to replacing it recently because i need a laptop with vulkan support but then it turned out that mesa actually…

> What on earth are you doing to your laptops that has you convinced that six years is an exceptionally long lifespan!?!? When I was a poor college student, I kept buying $500 Windows laptops and they kept breaking after 1-2 years.

Weren't they $800-900 a second ago?

Re: Apple has a memory problem and we're all paying for it

#138

Another version of what Bill Gates said: '640K software is all the memory anybody would ever need on a computer.'.

He probably never said that. https://www.computerworld.com/article/2534312/the--640k--quo...

Maybe not, but it's still a valid analogy to what Apple is doing now.

Re: Apple has a memory problem and we're all paying for it

#139
post #118

I think this is just about pricing. A Macbook with 8GB RAM is a shitty product. But you get to advertise that Macbooks start at 1599. Apple does a lot of analytics maybe they see that there's really a group of people who just use the default apps with some light browsing. But I think it's rather unlikely.

But why can't it start at $1600 with 16GB? For hundreds less you can get something like a Surface Laptop 5 or Asus Zenbook with similar builds and hi-res screens with 16GB RAM.

Re: Apple has a memory problem and we're all paying for it

#140
post #127

Earlier quoted context omitted.

Yes I have seen this. The disk write speed randomly gets super slow and tons of stuff just locks up at random as many many programs assume it'll be fine to do little writes synchronously. The whole machine becomes totally unusable. I see it on computers (and phones too) whose owners bought the smallest capacity to avoid spending money on something they didn't think they would use.

That's not what happens when an SSD reaches the write limit. It'll simply not let you write more data and becomes read only.

I have personally experienced this issue. It just gets very slow before slowly dying outright.
Post reply on HN