How Much Ram Do You Really Need in a MacBook?

16 gigs or 24? What's the right amount of RAM? There comes a point in every Mac computer purchase where the tech enthusiast must make a choice. Do I purchase the default 16 gigs of RAM or upgrade to 24? Or is it 32 GB? You mad man. First, I'd like to point out, in case you're unaware of current events and trends in computer hardware prices, we are in a literal RAM shortage crisis at the moment. And that's actually very real and not hyperbole. and more and more signs are pointing to it not being temporary. And I promise I'll spend less than a minute on this, but stick with me here. My PC, which I mainly just use to play Ano 1800 and ANO17, now has 64 gigs of RAM in it. I don't need that at all. But when I built that just a few months ago, this many sticks of RAM was just a $200 component. Here's the actual receipt. It was just $109 per 32 gigs of RAM. And this is even the fancy RGB RAM. at this moment to build that same computer. These four sticks with 64 gigs of RAM, this represents $1,100 of DDR5, it's completely bonkers. That is a lot. Apple has historically had annoyingly high RAM prices where upgrading from 16 to 32 gigs was a $400 purchase. That's still a lot, but something Apple did that Corsair seems not to have done, they signed long-term infrastructure deals with their suppliers for this very reason. When there's a shortage, an outage, a spike in price, Apple's RAM prices are locked in. So, where the bump from 16 to 24 was $200 6 months ago, it's still $200 today. That said, I'm also reading that this RAM situation is not a short-term problem. The sheer quantities of monsterized data centers that are being built just as quickly as they can right now as part of Earth's new AI backbone will continue to slurp up the majority of RAM modules that can be produced over the next few years. Apple's RAM price runway will run out and I don't know when. So, right now might be the very best time to pay the Apple tax as it's the cheapest DDR5 in town today, right? So, RAM costs a lot and I'm not here to talk you into buying more RAM than you need. I benefit exactly 0 or any internet credibility points based on the size of your purchase. So, let's talk about RAM first. Enough RAM is the goal. Having not enough RAM cripples your computer. And ever since Mac OS 10.9 Mavericks, also the name of my tiny little hybrid pickup truck outside, Mac OS has started using RAM compression with any unused application or Chrome tab, Grace, that's open in the background and not being actively used. If Mac OS decides that it thinks you're going to need more RAM in your current running program, it will compress the state of those background programs RAM amazingly sometimes up to as much as 60%. So, if you've got Chrome, Teams, and Mail open, but then also, I don't know, you were listening to some music, you were asking chat GPT some questions, and then you mapped out where you're going to go for lunch. So, you've got these six applications open, and they're using something like 8 gigs of RAM, and the system's using up another four. And then you go and open up Lightroom to do some editing of some raw photos you shot because, I don't know, maybe that's your job. Mac OS can package up the state of those 8 GB of working RAM into a chunk maybe about 3 GB large, freeing up almost 5 gigs of extra working RAM to use for Lightroom, so it can run at full speed without getting into the swap. Swap is the last line of defense after your computer has completely run out of working RAM and starts using your SSD as RAM. The RAM on these computers is a minimum of 25 times faster than your SSD can push data through. When this would happen with older computers that had spinning hard drives, everything just grinds to a halt cuz spinning hard drives are like hundreds of times slower than RAM. You could say thousands of times slower at random access. And that's usually, but not always, what's happening when you see your cursor turn into a little spinning beach ball still today. Fast forward to 5 years ago with the advent of Apple silicon. Mac OS switched from a Wilson Kaplan to an LZ dictionarybased compression algorithm to radically speed up the process of RAM compression. And I only say these two words to sound more authoritative on this subject. I did a lot of research and reading of articles for this one. Many hours in this chair. Anyway, as of M1, and I found several sources to back this up, so I trust it to be true. Apple's SoC has added hardware acceleration to its RAM compression algorithm to make it run something like 10 times faster. This is kind of a made-up figure because Apple's not public with the speed of how this works, but back in 10.9, we were on DDR3 RAM, which only had a maximum throughput of like 13 GB per second. Can you even imagine? And now with DDR5, the lowest trim base model M4. This thing's LP DDR5X has a throughput of 120 GB per second, 10 times faster. The M4 Max can push 564 GB per second through its RAM. That's the entire capacity of this computer's hard drive in 1 second. I'll call it a hard drive if I want to. That was a lot of words and numbers. This might be easier. RAM is 10 times faster now than when Apple started doing this. And hardware accelerated compression and decompression of your computer's RAM is nearly instant and uses almost no power. Which all boils down to a computer with 16 gigs of RAM can actually utilize something like 20 to 24 GB of RAM through compression before actually running out. And that's really how it works. and I can demonstrate it. I want to be thorough on what's what here with a fresh restart of this M5 MacBook Pro. And the M4 Air works exactly the same way. The RAM is a little faster on this guy, but not noticeably. Also, check out these numbers. 2026, we're going to hit 100,000. Going to get that little silver plaque. You're all going to be part of it. So, please subscribe if you're not subscribed already. Make that number bigger. So, with nothing but the system software running, and I made sure to go into login items and turn everything off, so there's nothing extra running. This computer's using almost 8 GB of RAM. So, let me go in and just start opening stuff. I just set up my normal workday. We get Teams going for chat and video calls. I'll open up mail, photos. Why not? I'll open up music. I always have pages running just in case an idea pops into my head. And then, let's get a bunch of tabs going in a browser. We need an operation schedule. Get a Reddit going. slash.news feed. I'll open up a bunch of videos from the YouTube suggestions. And hang on, I got to make sure these videos actually load up so they take up RAM. And this whole time, I've made sure to leave activity monitor visible on the bottom of the screen. And you can see RAM used actually stopped going up about halfway through all of that. But the compressed RAM is almost up to 5 GB. And for lightweight uses like all this stuff here, compressed RAM will not be any noticeably slower at all. I can actively be using 13 gigs of RAM and have another 5 GB of compressed RAM. nothing slows down. So, if I go back to the photos app, I can scrub through 36,000 photos with no lag whatsoever. I'll bet I can even open up something in AutoCAD and stay out of the swap at this point. And yep, you can see that took about two more gigs to run AutoCAD. But basically, the computer has a lot of different buckets full of RAM that it can move around before having to get into the dreaded swap page out system. And if you see this cached files number, we're going to get back to that in a bit. But first, let's keep going. Adobe Photoshop. Wab. Here's a thumbnail from a while ago. Ooh, we've hit the point of yellow memory pressure. This means the computer is just about out of options. Whatever can be compressed has been compressed. And here is the point where we are just about to run out of RAM for real. So yes, if you commonly go through your day with a dozen applications open and another dozen or more tabs open on your browser and this is just the way you want to live your life, then just pay that extra 200 bucks and get the 24 GB of RAM. Because as you can see, thanks to compression, you really have to have a lot of things open and running to find the bottom of 16 GB. Or it takes a single 8K video timeline running without proxies. And yes, it wouldn't be a Mac computer review without defaulting to Final Cut Pro to make some kind of point. The sheer amount of data in this 8K footage is so immense. There are 800 million pixels that need to be loaded and reconstructed from an H.265 265 video stream that just opening it and beginning playback starts you at about 13 GB of used RAM. And for videos with a second camera angle like mine, you can just forget about it if you want to edit 8K video without proxies on just 16 gigs of RAM. Or you can just edit with proxies. And in that case, you can have an 8K timeline and Photoshop and Lightroom. And I don't know why, but you still have some YouTube videos open. That's fine. Keep it up. No swap. And that's a pretty good argument right there that 16 gigs of RAM is still completely fine for anyone who's not working on large data sets or messing with LLMs running locally. And if you are doing that stuff, you already know what you need. So this information is not helpful for your descent. If you can do all this with 16 gigs of RAM, what do you even get out of getting more? Too much smoke. One of my New Year's resolutions is to learn how to actually use my haze machine. I mix up my own smoke juice. I make it too thick. It's my problem. When your computer's using RAM compression, it is fast enough that you can't even tell. And this is not something that wears out your RAM, unlike hitting the swap all day every day can wear out your SSD prematurely. SSDs do have a finite number of read and writes, so they will eventually go bad. Although, I still have one from 2013 that works fine. With more RAM comes more RAM cash, not to be confused with RAM cash, which is what you're going to need to build a new PC right now. I literally had to go to the bank and take out money for this bit. I don't use cash for anything anymore. RAM cache is your computer preloading programs and libraries and databases into its RAM. Stuff Mac OS thinks you're likely to open up again soon because maybe you use it a lot. Moving over to my more beefed up M4 Mac Studio. This guy has 36 gigs of RAM. So, this computer will be caching more RAM than my M4 Air even has. While this is running, there are 22 GB of RAM sitting in the cache, but none of it is compressed and no swap has been used. And this is the reason why, let's say you download and open Photoshop for the very first time, it'll take maybe 9 seconds of loading before you're actually in the app. And luckily, Photoshop just updated to version 27, so I had a chance to screen record this thing taking a long time. But if you tend to use Photoshop a lot and you have extra RAM on your computer, when you go back to open it up again in the future, it'll load in just about 2 seconds. And if you see that happening, it's because Mac OS noticed that you have a bunch of available RAM. Nothing in the immediate future looked like it was going to use that RAM. and I frequently go back and open up Photoshop. So, the Mac has gone ahead and just preloaded Photoshop and held it in the RAM just in case, which meant the next time I opened it, it was literally already open just in secret. This whole number, this 22 GB of cached RAM is totally available everywhere else. So, if Mac OS guessed wrong, it can dump those files it cached in an instant to use that RAM if I were to open up that 8K video timeline and suddenly needed a ton of it. This light shouldn't be facing me. That was for the chair spot. Oh, much more dramatic, much better lighting. But anytime I'm not using it, Mac OS is going to try to preempt programs and files and libraries and just all of the things that it thinks I'm most likely to use next. So, that's one way having extra RAM makes your computer feel more responsive. Programs won't run any faster with 36 gigs of RAM compared to 16 gigs of RAM if they don't require more than 16 gigs of RAM to run. But, they might open up much, much faster if you are a very repetitive computer user, and aren't we all? Then there are some programs that can just eat up however much RAM you can give to them. Blender, for instance, can load enormous scenes fully into RAM, and that program really does just get more and more responsive the more RAM you give it. And then there's those of you who run virtual machines. Other containers in the background, like Plex, for instance. If you just have a Plex server that's always running, it's going to use a little bit of RAM. It's going to cut into your RAM budget. Or with virtual machines, you can run Windows on a Mac and you can give Windows like give Parallels a specific chunk of its own RAM that can't be touched by the Mac. Again, you people already know what you need. I think though, if you're planning to run Parallels with Windows programs alongside of Mac programs, I want you to go for 24 gigs. Something I will say for sure, at the company I work for, the people that I work with who got the 16 GB upgrade for their M1 and M2 machines have absolutely no problems to report and still no reason to upgrade those now 5-year-old computers. I am the technical director at this event production company. So, computer problems land on me. And every single person who's come to me and complained that either AutoCAD is broken or when they're on Teams video calls and they try to share their screen, everything is too laggy and sometimes freezes. Those cases are always with the people who got the default 8 GB of RAM in their M1 or M2 computers. So, 8 GB is out. That is not a realistic RAM spec going into 2026. But 16 GB still works awesome for office jobs and creative work. That said, I do think that my official recommendation going into 2026 is changing and I say you should spring for the 24 GB and that's really only for future proofing. Most of you, me included, don't need 24 GB today. But like with Mac OS 26, how everything has that glassy, reflective, refractive quality with the windows and menus, that all takes GPU cycles and the RAM in the Mac is unified. It's shared. So any extra GPU thing takes a little extra RAM. And who knows what OS30 is going to look like. So having 8 gigs more is going to make it so that you can stretch your new computer another few years before you're going to need an upgrade. These are my thoughts on the matter of RAM. Happy New Year. Go buy a chair. Duncan's back to the traditional colors and that makes me very happy. Goodbye. Subscribe to the channel.

How much ram should you get with a new Macbook Pro or MacBook Air Computer? 16GB? 24GB? 32GB!? In this video I go over how well Ram Compression works on MacOS and talk about RAM Cache, and who needs how much for what reasons. If you want to bet on sports while supporting my channel use this link! Get (2) 100% Profit Boost Tokens on FanDuel Sportsbook! https://fndl.co/cdkedn5 THE CHAIR: 🎉 US/CA Deals: Up to 30% OFF: https://libernovo.com/products/libernovo-omni-dynamic-ergonomic-chair?ref=NICHOLASJOHNSON 🎁 Limited Free 6-Piece Gift Pack 16GB MacBook Pro: https://amzn.to/3Ln03ql 24GB MacBook Pro: https://amzn.to/4shre6q 16GB M4 Macbook Air: https://amzn.to/49tHSbE TG Fan Control affiliate link: https://www.tunabellysoftware.com?fpr=nicholas82 My External Drive that I edit from: https://amzn.to/49b4KM2 NVMe Drive: https://amzn.to/44CzTXd My New 8K camera! https://amzn.to/4qes6XJ If you just want to support my channel and have some Amazon Shopping to do - if you use this link I'll get a 2% kickback! https://amzn.to/49ep6nw 0:00 Intro 0:21 RAM Shortage 2025-2026 2:16 RAM Compression in MacOS 5:11 RAM Compression Demonstration 7:39 RAM used editing 8k in Final Cut Pro 8:29 LiberNovo Chair Integration its a nice chair! 11:36 Does More RAM Speed Up Your Computer? 14:19 My Thoughts on How Much RAM you Need