There were a number of exciting announcements from Apple at WWDC 2024, from macOS Sequoia to Apple Intelligence. However, a subtle addition to Xcode 16 — the development environment for Apple platforms, like iOS and macOS — is a feature called Predictive Code Completion. Unfortunately, if you bought into Apple’s claim that 8GB of unified memory was enough for base-model Apple silicon Macs, you won’t be able to use it. There’s a memory requirement for Predictive Code Completion in Xcode 16, and it’s the closest thing we’ll get from Apple to an admission that 8GB of memory isn’t really enough for a new Mac in 2024.
Opens chrome on a 8GB Mac. Sees lifespan of SSD being reduced by 50%. After 2-3 years of heavy usage SSD starts to get errors. Apple solution: buy a new one. No wonder they are 2nd/3rd wealthiest company on the planet.
buy a new one.
Buy a new SSD and swap out the old one?
…buy a new SSD, right??
SSD is soldered to the board. With only 8GB you’ll be using the swap partiton a lot so for anything exceeding 8GB of RAM you will be using the SSD as a slower “RAM” which will wear it’s lifespan down by constantly writing/reading into it’ s swap partition.
“tHATs nOT tRuE the aRCHiteCTuRe iS cOmPlETlY dIffErEnT!!!1!11!!ONEONE!!!” <— Apple fanboys when this was predicted on launch of the M1 🤖
No you don’t understand the architecture is different and so the laws of physics don’t apply. Constantly energizing and de-energizing capacitors can only increase life expectancy, everyone knows that.
The only people more cultish than Apple fans are Tesla/Elongated Muskrat fans.
You don’t get the most valuable company by selling a SSD. So, yeah a new Mac of course.
Well they do charge particularly hard for SSDs as well. They’ve found a way to eat the cake twice.
I think SSDs are also soldered to the mainboard on most apple products.
The Mac Studio uses a standard NVMe SSD but if you replace it with anything that you didn’t buy from Apple with a 500%+ markup, the new drive simply won’t work.
Oh, my sweet summer child…
Nah ur nat doin that with apple. Cmon just buy a new PC! Wa don car abt the env! Who cares anyway! Cmon not that expensive
deleted by creator
They moved to on-die RAM for a reason: To nickel and dime yo ass.
I needed to expense a Mac Mini for iOS development, and everyone (Me, the company, our purchasing department) was baffled at how much it cost to get 16 GB. And they only go up to 24GB. Imagine how much they’ll charge for 32 in a year!
It’s technically a bit faster, but yeah, I think charging more is the bigger motivation.
Companies primarily make decisions to maximise the profitability of someone and it’s never the consumer.
It’s a bit first but if their primary motivation was performance improvements they wouldn’t be soldering 16 GB.
If you’re going to weld shoes to your feet, you better at least make sure that they’re good shoes.
Why not? There is a performance benefit to being closer to the CPU, and soldering gets you a lot closer to the CPU. That’s a fact.
Yeah but if you’re only putting 8 GB of RAM on then you’re also going to be constantly querying the hard drive. So any performance gain you get from soldering, is lost by going all the way to the hard drive every 3 microseconds.
It’s only better performance on paper in reality there’s no real benefit. If you can run an application entirely entirely within the 8 GB of RAM, and assuming you’re not running anything else, then maybe you get better performance.
And that’s the idea. Soldering memory is an engineering decision. How much to solder is a marketing decision. Since users can’t easily add more, marketing can upsell on more RAM.
It’s not “on paper,” the RAM itself is performing better vs socketed RAM. Whether the system runs better depends on the configuration, as in, did you order enough RAM.
I can’t tell if you’re a stooge or if you really think that. I hope you are stooge, because otherwise that’s a really stupid position you’ve decided to take and you clearly don’t actually understand the issue.
I’m pretty sure I do understand the issue. Here are some facts (and an article to back it up):
- putting memory closer to the CPU improves performance due to less latency - from 96GB/s -> 200 (M1) or 400 (M1 Max) GB/s
- customers can’t easily solder on more RAM
- Apple’s RAM upgrades are way more expensive than socketed options on the market
And here’s my interpretation/guesses:
- marketing sees 1 & 2, and sees an opportunity to do more of 3
- marketing probably asked engineering what the bare minimum is, and they probably said 8GB (assuming web browsing and whatnot only), though 16GB is preferable (that’s what I’d answer)
- marketing sets the minimum @ 8GB, banking on most users who need more than the basics to buy more, or for users to buy another laptop sooner when they realize they ran out of RAM (getting after-sale RAM upgrades is expensive)
So:
- using soldered RAM is an engineering decision due to improved performance (double socketed RAM w/ Intel on M1, quadruple on M1 Max)
- limiting RAM to 8GB is a marketing decision
- if you don’t have enough RAM, that doesn’t mean the RAM isn’t performing well, it means you don’t have enough RAM
Using socketed RAM won’t fix performance issues related to running out of RAM, that issue is the same regardless. Only adding RAM will fix those performance issues, and Apple could just as easily make “special” RAM so you can’t buy socketed RAM on the regular market anyway (e.g. they’d need a different memory standard anyway due to Unified Memory).
I have hated Apple’s memory pricing for decades now, it has always been way more expensive to add RAM to an Apple device at order time vs PC competitors (I still add my own RAM to laptops, but it’s usually way cheaper through HP, Lenovo, etc than Apple at build-time). I’m not defending them here, I’m merely saying that the decision to use soldered RAM makes a lot of engineering sense, especially with the new Unified Memory architecture they’re using in the M-series devices.
Sounds like one of those rare cases where engineering and marketing might agree on something.
Mac Mini is meant to be sort of the starter desktop. For higher end uses, they want you on the Mac Studio, an iMac, or a Mac Pro.
I assumed that the Mini was the effectively a Mac without a monitor. Is it relatively underpowered too?
Its not underpowered for average users, but it’s not meant for professional uses beyond basic office work.
Similar to the mini they offer the Studio which doesn’t have a monitor built in https://www.apple.com/mac-mini/compare/?modelList=Mac-studio-2023,Mac-mini-M2
Then for the higher end uses they offer a more typical tower format https://www.apple.com/mac-pro/
But would an M1 Mini be similar to an M1 iMac?
As far as I understand, the Mac lineup don’t have screens, the IMacs are stationary and do have a screen, the MacBooks are the laptops.
Oh man, I remember so many people defended 8GB since the M1 first came out (and since).
I always argued it would significantly reduce the lifetimes of these machines if you bought one, not just because you’d be swapping a lot more on the (soldered in BTW) ssd, but because after a few years of updates it would become unbearably slow, or hardware would fail, or both.
Didn’t stop people constantly “tHe aRchITecTuRE iS cOmPlETelY diFFeRenT!!!”
Sure it’s different, but it’s still just a computer. A technical person can still look at the spec sheet and calculate effective performance accounting for bus widths etc.
Disclosure: I bought a top spec 16GB M1 Mac Air on launch and have been extremely happy with it - it’s still going strong.
Didn’t stop people constantly “tHe aRchITecTuRE iS cOmPlETelY diFFeRenT!!!”
Different Turing Machine on different math and alternative physics, I guess.
I bought a top spec 16GB M1 Mac Air on launch
My condolences.
EDIT: do people geuenly belive that math doesn’t apply to Apple’s products or they just don’t understand even such concentrated sarcasm?
Why do they struggle so much with some “obvious things” sometimes ? We wouldn’t have a type-C iphone if the EU didn’t pressured them to do make the switch
They don’t “struggle”. They are intentional and malicious decisions meant to drive revenue, as they have been since the beginning.
The E-Mac (looks like a toilet, sounds like a jet) came with 256 MB of RAM in one of the two slots, adding a 512MB stick was dirt cheap (everyone had at the very least 1GB on their PC), well it was dirt cheap except if you bought it from Apple…
It’s how Apple monetizes their customers. Figuring out an artificial shortcoming they can sell as an upgrade to them (check out dongles for example).
💸💸💸
They didn’t have a reason to switch to USB-C, and several reasons to avoid it for as long as possible. Their old Lightning connector (and the big 30-pin connector that came before it) was proprietary, and companies had to pay a royalty to Apple for every port and connector they manufactured. They made a lot of money off of the royalties.
I don’t consider that an “admission” at all…
To be fair there are only two reasons I hate it:
- People incorrectly use term UMA
- It’s crApple
On Linux if you don’t compile rust or firefox 8GB is fine. 4 is fine too.
I’d love to have 8GB of RAM. The SOC I’m working with has only 2K ;-)
Bet your compiler isnt running on that hardware either ;-)
Luckily, no ;-)
2 kilobytes?
Yes. 2 kilobytes. Coincidentally, this is as big as the displays internal buffer, so I cannot even keep a shadow copy of it in my RAM for the GUI.
Thanks for clarification.
I’ve never seen backbuffer called shadow copy.
And I have never heard it called “backbuffer”, so we are even.
I guess so.
Example: https://www.khronos.org/opengl/wiki/Default_Framebuffer#Double_buffering
EDIT: Wait. Do you have framebuffer at all? Because from sounds of it, you might not even have it at all. If you don’t store entire frame in RAM, then you don’t have framebuffer, not just backbuffer.
I never said anything about framebuffers. The 256x64 pixel display in 16 brightness levels probably has something comparable inside. I just tell it that i want to update a rectangle, and send it some data for that.
It should have.
Then, if you don’t store contents of entire screen in memory, which simple math says you can’t, I was partially wrong(depending on if you don’t count buffer in display as framebuffer) when interpreted “shadow copy” as backbuffer.
Come on, man, AVR chips aren’t SoCs except in the technical sense.
No AVR, it’s a small LPC from NXP. Chosen for the price, of course, but I have to somehow squeeze the software in it. At this point, even 8k would make me happy…
NXP, fancy. I expected ST, AVR, nRF, WCH or some chinese cheaptroller.
Why them? Something to do with NFC?
Man, microcontrollers are namegivers of SoC.
I have everything from apple but I know that 8gb is basically planned obsolescence in disguise.
We pay serious extra cash for just a ‚notch’ more refined experience. However I had to try to buy every possible thing from apple at least once in my life to see if it is worth it and basically only M4 iPad Pro 13 is truly worth the money and irreplaceable for me.
Everything else is nice for someone who is super lazy like me but can be easily replaced with not much difference for cheaper shit
Every time I compare specs to prices on Apples website, I get irrationally angry.
I have a macbook air m2 with 8gb of ram and I can even run ollama, never had ram problems, I don’t get all the hate
maybe in a browser using external resources. open some chrometabs to feel the pain. apple is a joke.
vscode + photoshop + illustrator + discord + arc + chrome + screen recording and still no lag
so not a single cool app and yet you own a computer
wtf does that mean
Which model with how many parameters du you use in ollama? With 8GB you should only be able to use the smallest models, which ist faaaar from ideal:
You should have at least 8 GB of RAM available to run the 7B models, 16 GB to run the 13B models, and 32 GB to run the 33B models.
llama3:8b, I know it’s “far from ideal” but only really specific use cases require more advanced models to run locally, if you do software development, graphic design or video editing 8gb is enough
edit: just tried it after some time and it works better than I remembered showcase
Removed by mod
I wonder what is the general use for the Mac Mini, MacBook Air, iMac, and MacBook Pro? People generally seem to do all the lightweight stuff like social media consumption on their phones; and desktops/laptops are used for the more heavy-weight stuff. The only reason I’ve ever used a Mac was for IOS development.
I have a friend who is self-employed. He uses an iPhone and a MacBook Air. He only uses iMessage, Numbers, Safari and Apple Music for entertainment. He gets away with 8gb just fine and rarely has to reboot.
He probably could use a Chromebook or something even lighter, but the support and ecosystem were enough for him to pay the premium. His time is valuable to him so it was worth it to him.
Any computer with only 1MB of RAM is also usable … as paperweight.
Well, 2000€ for a “Pro” model of the Macbook 14" with only 8GB RAM sounds a bit strange, tbf. And +230€ for +8GB is straight up greedy.
They said “Actually, 8GB on an M3 MacBook Pro is probably analogous to 16GB on other systems” and well , that’s definitely not the case for their upcoming AI usecases, because - and many people seem to overlook that - their RAM is shared RAM (or as they call it “unified memory”), which means that the GPU is limited by these 8GB of (V)RAM because it can only use what is left by the System usage.
Sorry, boo, everyone wants to hate Apple these days. It’s the Zeitgeist. Even if you say something reasonable or perhaps factual, the people are against you and will react violently.
Accurate.
Removed by mod
The unreasonable escalation of your response really makes you come across as exceedingly insecure.
I was quite literally illustrating the absurdity by being similarly absurd. Telling people to shut the fuck up about an issue is funny as hell to respond with a similar statement.
See the OP’s ending sentence for reference.
If you choose to be a weak little quiet corporate Stan then that’s up to you. Apple is well aware that third party apps exist and they’re well aware that machines with less ram will need replaced far sooner than machines with more. RAM is cheap and Apples intigrated memory is no different in the regard. The only reason to use less is planned obsolescence. If you don’t believe that then you’re either Tim Cook or you’re an idiot.
What is the obsession with shitting on people’s choices? I don’t understand the irony of demanding choice in this industry, then shitting on people when they make a choice you don’t agree with.
No one is shitting on other people’s choices. They are criticizing a major corporations choices to skimp on specs while charging a premium price. Specs that can’t be upgraded and will absolutely lead to a shorter usable life. I find it odd that people get upset about criticism that isn’t aimed at them at all. The only thing I can think is maybe they realize they were ripped off after putting so much money into Apple products and they need to defend their financial decisions. Even then I don’t fully understand. I’ve purchased overpriced junk many times and don’t feel the need to defend the offending company. Maybe it’s because Apple has managed to make their customers feel like they’re in an exclusive club even though everyone uses Apple products these days. A publicly traded company is around to make money and nothing more. They should never be held in reverence.
What is the obsession with shitting on people’s choices?
As much as people want to act like they are better than they were, say, 100 years ago, it’s not really true. Humans are really just advanced monkeys running around and very few can actually surpass that nature.
This we can agree on.
And now all the fan boys and girls will go out and buy another MacBook. That’s planned obsolescence for ya
And the apple haters will keep making this exact same comment on every post using their 3rd laptop in ten years while I’m still using my 2014 MacBook daily with no issues.
Be more original.
Nice attempt to justify planned obsolescence. To think apple hasn’t done this time and time again, you’d have to be a fool
👍
-posted from my ten year old MacBook which shows no need for replacement
And is what, 3 or 4 operating systems behind due to it being obsolete
At which point did Apple decide your MacBook was too old to be usable and stop giving updates or allow new software to run on it?
Still gets security updates. All the software I need to run on it runs on it.
My email, desktop, and calendar all still sync with my newer desktop. I can still play StarCraft. I can join zoom meetings while running Roll 20. I can even run Premiere and do video editing… to a point.
I guess if you need the latest and greatest then you might have a point, but I don’t.
This whole thread is bitching about software bloat and Apple does that to stop the software bloat on older machines, but noooo that’s planned obsolescence. 🙄
They will keep making the same comment as long as it keeps being true.
- Typed from my 2009 ThinkPad
Meanwhile your 2014 MacBook stopped receiving OS updates 3 years ago.
Weren’t you just complaining about software bloat?
…no?
This is pretty much it. People really just want to find reasons to hate Apple over the past 2 - 3 years. You’re right, though, your Mac can run easily for 10+ years. You’re good basically until the web browsers no longer support your OS version, which is more in the 12-15 year range.
In fairness, most computers built after around 2014-2016+ last way longer, performance started to level off not long after that. After all, devs write software for what people have, if everyone had 128 gigs of RAM we’d load everything we could think of into memory and you’d need it to keep up
Macs did have some incredible build quality though, the newer ones aren’t holding up even close to as well. I’m still using a couple 2012 Macs to play videos, it’s slow as hell when you interact, but once the video is playing it still looks and sounds good
I still have a fully functioning Windows 95 machine.
My daily driver desktop is also from around 2014.
That’s pretty sick actually
These were obsolete the minute they were made, though… So it’s not really planned obsolescence. I got one for free (MacBook Air), and it’s always been trash.
I have an M2 MBA and it’s the best laptop I’ve ever owned or used, second to the M3 Max MBP I get to use for work. Silent, battery lasts all week, interface is fast and runs all my dev tools like a charm. Zero issues with the device.
Someone who is buying a MacBook with the minimum specs probably isn’t the same person that’s going to run out and buy another one to get one specific feature in Xcode. Not trying to defend Apple here, but if you were a developer who would care about this, you probably would have paid for the upgrade when you bought it in the first place (or couldn’t afford it then or now).
Well no, not this specific scenario, because of course devs will generally buy machines with more RAM.
But there are definitely people who will buy an 8GB Apple laptop, run into performance issues, then think “oh I must need to buy a new MacBook”.
If Apple didn’t purposely manufacture ewaste-tier 8GB laptops, that would be minimised.
I wouldn’t be so sure. I feel like many people would not buy another MacBook if it were to feel a lot slower after just a few years.
This feels like short term gains vs. long term reputation.
And why they solder the RAM, or even worse make it part of the SoC.
There are real world performance benefits to ram being as close as possible to the CPU, so it’s not entirely without merit. But that’s what CAMM modules are for.
But do those benefits outweigh doubling or tripling the amount of RAM by simply inserting another stick that you can buy for dozens of dollars?
Yes, there are massive advantages. It’s basically what makes unified memory possible on modern Macs. Especially with all the interest in AI nowadays, you really don’t want a machine with a discrete GPU/VRAM, a discrete NPU, etc.
Take for example a modern high-end PC with an RTX 4090. Those only have 24GB VRAM and that VRAM is only accessible through the (relatively slow) PCIe bus. AI models can get really big, and 24GB can be too little for the bigger models. You can spec an M2 Ultra with 192GB RAM and almost all of it is accessible by the GPU directly. Even better, the GPU can access that without any need for copying data back and forth over the PCIe bus, so literally 0 overhead.
The advantages of this multiply when you have more dedicated silicon. For example: if you have an NPU, that can use the same memory pool and access the same shared data as the CPU and GPU with no overhead. The M series also have dedicated video encoder/decoder hardware, which again can access the unified memory with zero overhead.
For example: you could have an application that replaces the background on a video using AI. It takes a video, decompresses it using the video decoder , the decompressed video frames are immediately available to all other components. The GPU can then be used to pre-process the frames, the NPU can use the processed frames as input to some AI model and generate a new frame and the video encoder can immediately access that result and compress it into a new video file.
The overhead of just copying data for such an operation on a system with non-unified memory would be huge. That’s why I think that the AI revolution is going to be one of the driving factors in killing systems with non-unified memory architectures, at least for end-user devices.
I feel like this is an arguement for new specialized computers at best. At worst it shows that this AI crap is even more harmful to the end user.
That’s a fantastic explanation! Thank you!
Bus goes Vrrrroom vrrooom. Fuck AI.
And even if the out-of-the-box RAM is soldered to the machine, it should still be possible to add supplementary RAM that isn’t soldered for when the system demands it. Other computers have worked like this in the past with chip RAM but a socket to add more.
It’s highly dependent on the application.
For instance, I could absolutely see having certain models with LPCAMM expandability as a great move for Apple, particularly in the pro segment, so they’re not capped by whatever they can cram into their monolithic SoCs. But for most consumer (that is, non-engineer/non-developer users) applications, I don’t see them making it expandable.
Or more succinctly: they should absolutely put LPCAMM in the next generation of MBPs, in my opinion.
That’s extremely dependent on the use case, but in my opinion, generally no. However CAMM has been released as an official JEDEC interface and does a good job at being a middle ground between repairability and speed.
It’s an officially recognized spec, so Apple will ignore it as long as they can. Until they can find a way to make money from it or spin marketing as if it’s some miraculous new invention of theirs, for something that should just be how it’s done.
Parts pairing will do. That’s what Apple known for, knee capping consumer rights.
Apple’s SoC long predates CAMM.
Dell first showed off CAMM in 2022, and it only became JEDEC standardised in December 2023.
That said, if Dell can create a really good memory standard and get JEDEC to make it an industry standard, so can Apple. They just chose not to.
In this particular case the RAM is part of the chip as an attempt to squeeze more performance. Nowadays, processors have become too fast but it’s useless if the rest of the components don’t catch up. The traditional memory architecture has become a bottleneck the same way HDDs were before the introduction of SSDs.
You’ll see this same trend extend to Windows laptops as they shift to Snapdragon processors too.
People do like to downplay this, but SoC is the future. There’s no way to get performance over a system bus anymore.
There is. It’s called CAMM.
Funny that within one minute, they state the exact same problem.
If you actually watch past the first minute of the video, they explain that LPCAMM solves that problem…
deleted by creator
BUT BUT you’ll get 5% fasTEr SpeED!!! And MOrE seCuRiTy!!!
Well. The claim they made still holds true, despit how I dislike this design choice. It is faster, and more secure (though attacks on NAND chips are hard and require high skill levels that most attacker won’t posses).
And add one more: it saves power when using LPDDR5 rather DDR5. To a laptop that battery life matters a lot, I agree that’s important. However, I have no idea how much standby or active time it gain by using LPDDR5.
This isn’t a big deal.
If you’re developing in Xcode, you did not buy an 8GB Mac in the last 10-years.
If you are just using your Mac for Facebook and email, I don’t think you know what RAM is.
If you know what RAM is, and you bought an 8GB Mac in the last 10-years, then you are likely self-aware of your limited demands and/or made an informed compromise.
Funny: knowing that you only get one shot, I bought 32GB of RAM for my Mac Mini like 1.5 years ago. I figured that it gave me the best shot of keeping it usable past 5 years.
If you know what RAM is, and you bought an 8GB Mac in the last 10-years, then you are likely self-aware of your limited demands and/or made an informed compromise.
Or you simply refuse to pay $200+ to get a proper machine. Like seriously, 8GB Mac’s should have disappeared long ago, but nope, Apple stick to them with their planned obsolescence tactics on their hardware, and stubbornly refusing to admit that in 2023 releasing a MacBook with soldered 8Gb of RAM is wholy inadequate.
Yeah, the 8GB model’s purpose is to make an “starting at $xxxx” price tag possible.
I get around this by simply not buying a Mac. Free’s up so much money for ram.
deleted by creator
I’m not gonna stand up and declare that 8gb is absolutely fine, because in very short order it won’t be. But yeah, currently for an average use case, it is.
My work Mac mini has 8gb. It’s a 2014 so can’t be upgraded, but for the tasks I ask of it it’s ok. Sure, it gets sluggish if I’m using the Win11 VM I sometimes need, but generally I don’t really have any issues doing regular office tasks.
That said, I sometimes gets a bee in my bonnet about it, so open Activity Monitor to see what’s it’s doing, and am shocked by how much RAM some websites consume in open tabs in Safari.
8gb is generally ok on low end gear, but devs are working very hard to ensure that it’s not.
imagine showing this post to someone in 1995
shit has gotten too bloated these days. i mean even in my head 8GB still sounds like ‘a lot’ of RAM and 16GB feels extravagant
Absolutely.
Bad, rushed software that wires together 200 different giant libraries just to use a fraction of them and then run it in a sandboxed container with three daemons it needs for some reason doesn’t mean “8 Gb isn’t enough”, it means write tighter, better software.
That ship has long sailed unfortunately. The industry gave up on optimization in favour of praying that hardware advancements can keep up with the bloat.
You just have to watch your favorite tablet get slower year after year to understand that a lot of this is artificial. They could make applications that don’t need those resources but would never do so.
Guy from '95: “I bet it’s lightning fast though…”
No dude. It peaks pretty soon. In my time, Microsoft is touting a chat program that starts in under 10 seconds. And they’re genuinely proud of it.
And latency is more shit than it ever was
We measure success by how many GB’s we have consumed when the only keys depressed from power on to desktop is our password. This shit right here is the real issue.
I once went for lower CAS timing 2x 128MB ram sticks (256 MB) instead of 2x 256s with slower speeds because I thought 512MB was insane overkill. Realized how wrong I was when trying to play Star Wars galaxies mmorpg when a lot of people were on the screen it started swapping to disk. Look up the specs for an IBM Aptiva, first computer my parents bought, and you’ll understand how 512MB can seem like a lot.
Now my current computer has 64 GB (most gaming computers go for 32GB) at the time I built it. My workstation at work has 128GB which really isn’t even enough for some workloads we have that use a lot of in-memory cache… And large servers can have multiple TB of RAM. My mind has been blown multiple times.
I still can’t fully accept that 1GB is not normal, 2GB is not very good, and 4GB is not all you ever gonna need.
If only it got bloated for some good reasons.
I remember when I got my first computer with 1GB of RAM, where my previous computer had 64MB, later upgraded to 192MB. And there were only like 3 or 4 years in between them.
It was like: holy shit, now I can put all the things in RAM. I will never run out.
The moment you use a file that is bigger than 1GB, that computer will explode.
Some of us do more than just browse Lemmy.
Wow. Have you ever considered how people were working with files bigger than total RAM they had in the normal days of computing?
So in your opinion if you have 2GB+ of a log file, editing it you should have 2GB RAM occupied?
I just have no words, the ignorance.
High quality content is the reason. Sit in a terminal and your memory usage will be low.
So we’re just going to ignore stuff like Electron, unoptimized assets, etc… Basically every other known problem… Yeah let’s just ignore all that
Is Electron that bad? Really? I have Slack open right now with two servers and it takes around 350MB of RAM. Not that bad, considering that every other colleague thinks that posting dumb shit GIFs into work chats is cool. That’s definitely nowhere close to Firefox, Chrome and WebStorm eating multiple gigs each.
Yes, it really is that bad. 350 MBs of RAM for something that could otherwise have taken less than 100? That isn’t bad to you? And also, it’s not just RAM. It’s every resource, including CPU, which is especially bad with Electron.
I don’t really mind Electron myself because I have enough resources. But pretending the lack of optimization isn’t a real problem is just not right.
First of all, 350MB is a drop in a bucket. But what’s more important is performance, because it affects things like power consumption, carbon emissions, etc. I’d rather see Slack “eating” one gig of RAM and running smoothly on a single E core below boost clocks with pretty much zero CPU use. That’s the whole point of having fast memory - so you can cache and pre-render as much as possible and leave it rest statically in memory.
CPU usage is famously terrible with Electron, which i also pointed out in the comment you’re replying to. But yes, having multiple chromium instances running for each “app” is terrible
First of all, 350MB is a drop in a bucket
People don’t run just a single app in their machines. If we triple ram usage of several apps, it results in a massive increase. That’s how bloat happens, it’s a cumulative increase on everything. If we analyze single cases, we could say that they’re not that bad individually, but the end result is the necessity for a constant and fast increase in hardware resources.
Just wanted to point out that the number 1 performance blocker in the CPU is memory. In the general case, if you’re wasting memory, you’re wasting CPU. These two things really cannot be talked about in isolation.
When (according to about:unloads) my average firefox tab is 70-230MB depending on what it is and how old the tab is (youtube tabs for example bloat up the longer they are open), a chat app using over 350 is a pretty big deal
just checked, my firefox is using 4.5gb of RAM, while telegram is using 2.3, while minimized to the system tray, granted Telegram doesnt use electron, but this is a trend across lots of programs and Electron is a big enough offender I avoid apps using it. When I get off shift I can launch discord and check it too, but it is usually bad enough I close it entirely when not in use
It sure is. I’m running ferdium at this very moment with 3 chat apps open, and it consumes almost a gigabyte for something that could take just a few megabytes.
What’s wrong with using Gifs in work chat lmao, can laugh or smile while hating your job like the rest of us.
Get a better job.
256MB or 512MB was fine for high-quality content in 2002, what was that then.
Suppose the amount of pixels and everything quadrupled - OK, then 2GB it is.
But 4GB being not enough? Do you realize what 4GB is?
They didn’t just quadruple. They’re orders of magnitude higher these days. So content is a real thing.
But that’s not what’s actually being discussed here, memory usage these days is much more of a problem caused by bad practices rather than just content.
I know. BTW, if something is done in an order of magnitude less efficient way than it could and it did, one might consider it a result of intentional policy aimed at neutering development. Just not clear whose. There are fewer corporations affecting this than big governments, and those are capable of reaching consensus from time to time. So not a conspiracy theory.
One frame for a 4K monitor takes 33MB of memory. You need three of them for triple buffering used back in 2002, so half of your 256MB went to simply displaying a bloody UI. But there’s more! Today we’re using viewport composition, so the more apps you run, the more memory you need just to display the UI. Now this is what OS will use to render the final result, but your app will use additional memory for high res icons, fonts, photos, videos, etc. 4GB today is nothing.
I can tell you an anecdote. My partner was making a set of photo collages, about 7 art works to be printed in large format (think 5m+ per side). So 7 photo collages with source material saved on an external drive took 500 gigs. Tell me more about 256MB, lol.
Yes, you wouldn’t have 4K in 2002.
4GB today is nothing.
My normal usage would be kinda strained with it, but possible.
$ free -h total used free shared buff/cache available Mem: 17Gi 3,1Gi 11Gi 322Mi 3,0Gi 14Gi Swap: 2,0Gi 0B 2,0Gi $
I can do a cold boot and show you empty RAM as well. So fucking what?
It’s not a cold boot and it’s not empty.
I chalk it up to lazy rushed development. Good code is art.
That’s not true at all. The code doesn’t take much space. The content does. Your high quality high res photos, 4K HDR videos, lossless 96kHz audio, etc.
But there are lots of shortcuts now. Asset packs and coding environments that come bundled with all kinds of things you don’t need. People import packages that consume a lot of space to use one tiny piece of it.
To be clear, I’m not talking about videos and images. You’d have these either way.
All these packages don’t take much memory. Also tree shaking is a thing. For example, one of the projects I currently work on has over 5 gigs of dependencies, but once I compile it for production, the whole code based is mere 3 megs and that’s including inlined styles and icons. The code itself is pretty much non-existent.
On the other hand I have 100KB of text translations just for the English language alone. Because there’s shit loads of text. And over 100MB of images, which are part of the build. And then there’s a remote storage with gigabytes of documents.
Even if I double the code base by copy pasting it will be a drop in a bucket.
I have a VPS that uses 1GB of RAM, it has 6-7 apps running in docker containers which isn’t the most ram efficient method of running apps.
A light OS really helps, plus the most used app that uses a lot of RAM actually reduce their consumption if needed, but use more when memory is free, the web browser. On one computer I have chrome running with some hundreds of MB used, instead of the usual GBs because RAM is running out.
So it appears that memory is full,but you can actually have a bit more memory available that is “hidden”
Same here. When idle, the apps basically consume nothing. If they are just a webserver that calls to some PHP script, it basically takes no RAM at all when idle, and some RAM when actually used.
Websites and phone apps are such an unoptimized pieces if garbage that they are the sole reason for high RAM requirements. Also lots of background bloatware.
This is resource reservation, it happens at an OS level. If chrome is using what appears to be alot of ram, it will be freed up once either the OS or another application requires it.
It just exists so that an application knows that if it needs that resource it can use X amount for now.
You can always switch to a text based terminal and free up your memory. Just don’t compain that YouTube doesn’t play 4K videos anymore.
Just don’t compain that YouTube doesn’t play 4K videos anymore.
strange, mpv handles it just fine
MPV doesn’t work in terminal (well, technically it does, but what’s the point of 4K HDR video in ASCII mode?). Please don’t confuse terminal emulator in GUI mode with a real text mode terminal.
The point is that your example use case of “YouTube 4k videos” doesn’t need a browser full of bloated js garbage.
The point is that MPV will use shitloads of memory too.
Actually lot less than the browser. Under 300MB, I just checked, and that’s mostly just the network buffer which is 150MB by default.
That’s about what my Slack is using, while being written in Electron, lol. Oh, you people…
KMSDRM is in terminal enough for me. Fbcon too.
EDIT: obviously not dummy terminal over UART or like that.
8GB of dedicated VRAM is hardly enough these days…
This is my biggest lament about getting a 2060 without knowing how important vram is. I can make it perform better and more efficiently a bunch of different ways, but to my knowledge, I can’t get around the 6GB vram wall.
Especially with 4k
4k rendering or 4k textures?