And people were really trying to convince me in a previous thread that Apple's magical smart memory access tech meant RAM capacity was now irrelevant 
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Video Demos Performance Differences Between 8GB and 16GB Apple M1 MacBook Pro
- Thread starter MacRumors
- Start date
- Sort by reaction score
SSD write speeds dictate how fast you can download something. Faster SSD write = greater capability to exploit a fast internet connection
When he just had the Max Yuryev channel I used his affiliate link and bought my Sony Mirrorless camera through him. Last year he's been really on fire with the Apple reviews. He's honest too because Apple doesn't send him any machines. He buys them all for his reviews. I hope Apple recognizes him. His brother is awesome too!.
Yes, this is a tech site, we all speak tech here.
The comment originated from the ABSURD idea that resale values are *ever* going to take that into consideration.
I think also Big Sur has much better memory management overall, not just on M1, but also on Intel. My old MBP also have 8GB of RAM, and are able to run more apps than before at the same time.
The present comparison of 8 vs. 16 leaves much to be desired. See my comments above.
I keep seeing these dire predictions, but that shoe hasn't dropped yet.
"Sure, it may be good in single core, but multithreaded applications will cripple it"
(swing and a miss)
"Sustained usage will cause thermal throttling like crazy"
(strike two)
"Applications that rely on the video card will just crawl"
(strike three, someone should be out)
"16 GB of memory isn't enough to open multiple applications"
(goes to app folder, cmd-a, cmd-o and strike four)
"16GB memory can't handle 4k60"
(8GB MB Air handles 4k120, better than an i9 32GB - ouch, that one hurts me personally)
"No gaming on that toy"
(play games better than Intel equivalent)
"Large music projects will kill it"
(base mini opens 64 virtual tracks + 75 tracks of virtual drummers + 256 instruments and over 900 plugins while playing smoothly and not maxing out the CPU)
So I will have to see what the swap file predictions yield. Someone will test it soon enough.
I am sure that one day, one of these predictions will be right because it isn't like Apple is infallible or that the M1 is perfect, but it certainly seems to be a hell of a lot better than anyone expects.
Yep, pretty sure we got that after your second post here. Cheers.
I have 1667 tabs open at the moment. Not for a particularly good reason, but because I'll see links on pages and think 'oh yeah, I'll have to check that out', middle click to open it in a background tab, and carry on with the page I was on. Then I'll either say 'well I should leave this tab open for reference because this bit is possibly important' or to otherwise establish a "big picture" of a particular problem I'm trying to solve (where I can quickly cycle through several tabs to piece something together).
You then run into the Wikipedia effect where you open a tab on a completely different topic and then branch out from there, abandoning the previous topic in other tabs. Weeks later you find that you've made a serious mistake and you start closing tabs... but oh... that looks interesting *middle click*.
It's a vicious cycle.
That being said, they're all suspended tabs. Adding up all of the helper processes, Firefox is barely using 2 GB at the moment. So I don't see the behavior as a real technical issue, just uh... a behavioral issue.
The thing is, often times you end up not upgrading when you expected because of several factors, at least in my experience. So saying I’ll keep this device for the next two years, it doesn’t always work like that.
MS Office 365 runs faster on an M1 through Rosetta 2 emulation than natively on MacOS x86. It's crazy.
Just the one browser.
Desktops like spaces? I never got the hang of using those. I'm a CMD-tabber. Although recently I got an ultrawide monitor, so now I have Firefox always open on the left and iTerm2 is usually on the right (sometimes replaced with VSCode or Xcode). Occasionally Signal will float in the middle.
As others have kind of noted free memory is not a great indicator of spare capacity in this scenario as this is a metric that any vanilla unix type system typically tries to minimize as much as possible. Free unused memory is wasted memory - much better to buffer IO with it so its caching something that you won't have to wait later on to get from a slower device such as an SSD. The system can and will give you all that cache space back if you need it at some point. So free memory is rather an indicator of a system's failure to cache effectively than it is an indication of super abundant capacity. Add in the aggressive swap/sleep behavior of the hybrid iOS/OSX memory management and I really don't know that these are useful memory stress tests.
I don't think so at all, unless someone can demonstrate that the SSD is prevalent point of failure due to wear and tear compared to other components. Keyboards wear out. Batteries wear out. I expect they would approach end of life earlier than the SSD in most cases.
Also keep in mind that people buy used machines for budget reasons. Extra ram might get you $50-100 extra used value a couple years later on a $200 upgrade. It's not going to reduce the overall cost of the machine on your end.
The problem with this is you are thinking about the M1 in terms of an x86 even though it is clear that the way it handles memory is totally alien to the x86 method or even Android's version of ARM.
What if you opened around 1000 Safari tabs first then ran the tests?
Ok so it would be x86 Marketing Department rather then Intel Marketing Department.
While the M1 architecture is periphery its instruction set (ARM) is not. Never mind that Swift is OpenSource so writing crossplatform/OS software is easy. Unless Parallels and VMWare drop the ball they should have VMs that performs better then the effort Microsoft tried.
All you kiddos who spent a lot of time overthinking and ended up future proofing your 15 inch and 16 inch MBP RIP. Suckers
You spend your days looking at an ocean of random internet info? How do you even keep track of tabs with 2000 open?
I think nearly everyone missed the point of the referenced video - it is not the speed of SSD swapping but how often the swapping occurs and the size of the swapped memory.
"An electric effect results from the fact that data can only be written on a storage cell inside the chips between approximately 3,000 and 100,000 times during its lifetime." - How long do SSDs last? For a given TBW larger actual RAM means less swapping and, therefore, a longer life for the SSD.