I know it’s not a like for like comparison, but the Pixel 9 Pro that launched this month has 16gb of RAM.
It’s a good comparison actually because Apple keeps saying that their ram is faster because it’s soldered (Which is true but only if you squint). I don’t really think it makes a difference because if you run out of space you still run out of space, the fact that you can access the limited space more quickly doesn’t really help.
Well phone RAM also tends to be solded onto the board too so it’s a pretty good comparison.
I remember back in the early 2000s when I saw a PDA with a 232mhz cpu and 64mb ram, and I realized how far technology had come since I got my computer with a 233mhz cpu and 64mb ram…
Obviously different architechtures, but damn that felt strange…
Some phones have 24gb since 2 years ago already
Which phone(s) had 24Gb standard?
I didn’t think any of those are the base model. Anything with Pro or Ultra in the name should have more than 8Gb of RAM in my opinion. It also seems dominated by OnePlus as the others listed are not really players in the larger market. You could possibly argue that Xaomi is but I’ve never even seen one of these phones in the real world. In fact it looks like most of these are only available in China variant.
I am writing this using my Xiaomi poco x3 pro, although it have 8gb ram and 256gb memory, it also have headphone jack and micro SD slot
Yup, while the current iPhone 15 Pro is the only model which has 8 GB of RAM, with the regular iPhone 15 having 6 GB. All iPhone 16 models (launching next month) will still only have 8 GB according to rumors, which happens to be the bare minimum required to run Apple Intelligence.
Giving the new models only 8 GB seems a bit shortsighted and will likely mean that more complex AI models in future iOS versions won’t run on these devices. It could also mean that these devices won’t be able to keep a lot of apps ready in the background if running an AI model in-between.
16 GB is proper future-proofing on Google’s part (unless they lock new software features behind newer models anyway down the road), and Apple will likely only gradually increase memory on their devices.
I don’t use Apple computers but if we’re going into phones, iOS is extremely memory efficient. I’m on a six year old XS max with 4GB and it works like the day I got it, running circles around Android phones half its age.
If you were being cynical, you could say it was planned obsolescence and that when the new ai feature set rolls out that you have to get the new phone for them.
I think they got caught with their pants down when everybody started doing AI and they were like “hey, we have this cool VR headset”. Otherwise they would’ve at least prepared the regular iPhone 15 (6 GB) to be ready for Apple Intelligence. Every (Apple Silicon) device with 8 GB or more get Apple Intelligence, so M1 iPads from 2021 get it as well for example, even though the M1’s NPU is much weaker than some of the NPUs in unsupported devices with less RAM.
They are launching their AI (or at least everything under the “Apple Intelligence” umbrella) with iOS 18.1 which won’t even release with the launch of the new iPhones, and it’ll be US only (or at least English only) with several of the features announced at WWDC still missing/coming later and it’s unclear how they proceed in the EU.
With how polished Apples AI on mobile was at launch compared to Gemini on Android at launch were it could not even do basics like timers I suspect Apple had it in the works for far longer and it would not have been a total surprise.
Also you are describing the situation at launch for new hardware, the software will evolve every year going forward and the requirements will likely increase every year. If I am buying a flagship phone right now I want it to last at least 3 years of updates, if not 5 years. The phone has to be able to cope with what is a very basic requirement that is enough RAM.
This isn’t some NPU thing, this is just basic common sense that more RAM is better for this, something the flagship iPhones could have benefited from for a while now.
I’m not sure if you’re agreeing or disagreeing with me here. Either way, hardware has a substantially longer turnaround time compared to software. The iPhone 15 would’ve been in development years before release (I’m assuming they’re developing multiple generations in parallel, which is very likely the case) and keep in mind that the internals are basically identical to the iPhone 14 Pro, featuring the same SoC.
AI and maybe AAA games like Resident Evil aside, 6 GB seems to work very well on iPhones. If I had a Pixel 6/7/8 Pro with 12 GB and an iPhone 12/13/14 Pro (or 15) with 6 GB, I likely wouldn’t notice the difference unless I specifically counted the number of recent applications I could reopen without them reloading. 6 GB keeps plenty of recent apps in memory on iOS.
But I’m not sure going with 8 GB in the new models knowing that AI is a thing and the minimum requirement for their first series of models is 8 GB is too reassuring. I’m sure these devices will get 5-8 years of software updates, but new AI features might be reduced or not present at all on these models then.
When talking about “AI” in this context I’m talking about everything new under the “Apple Intelligence” umbrella, like LLMs and image generators. They’ve done what you’d call “AI” nowadays for many years on their devices, like photo analysis, computational photography, voice isolation, “Siri” recommendations etc.
I was under the impression that ios used sleight of hand with apps to reduce memory footprint for inactive apps rather than how android manages its recent apps list? Is it still requiring special permissions to run non apple apps in the background as active tasks? AI will need to run the background and will need a decent chunk of RAM to do so.
I completely agree that changing the processor or revising NPU or similar is too much to do late stage, I reject that for increasing RAM or storage, both can be changed closer than 12 months from release and I would also reject that apple had the AI changes planned for much less than 12 months out as well. It just feels like a big fuck you to anybody buying a flagship from apple this year as it wont last the length of time it should do for normal consumers who would expect all of the latest AI features to roll out during the supported window.
I would say it is more so they can advertise a lower price. But then expect you to get the more expensive ones as the bare minimum is just not enough.
For the base model yeah, but apple loves charging a packet for more memory so I don’t see it for the top of the range models. Would be typical for them to only offer 16gb with the increased storage as well, just to bump the price up
Pretty much what NVIDIA is doing with their GPUs. Refusing to provide adequate future proof amount of VRAM on their cards. That’s planned obsolescence in action.
And like Apple, Nvidia has no shortage of fanboys that insist the pitiful amounts of (V)RAM is enough. The marketing sway those two companies have is incredible.
It’s a complete joke that Sapphire had an 8GB version of the R9 290X, what, 11 years ago or something? And yet Nvidia is still selling 8GB cards now, for exorbitant prices.
This happens if you sell your hardware as DRM key to use their software (i(Pad)OS, macOS etc. and Cuda)
The current GPU situation actually has me curious about AMDs upcoming Halo APU chips. They’re likely going to be pretty expensive relative to their potential GPU equivelent performance but if they work out similar to the combined price of a CPU and GPU then it might be worthwhile as they use onboard RAM as their VRAM. Probably a crazy idea but one I look forward to theory-building in spring when they release.
Welcome to 2010
I upgraded from 8GB to 16GB like 2 months ago.
Welcome to 2010 to you as well then!
Check out 9gag.com for funny pictures!
Or iFunny if you’re feelin’ saucy!
Good, 8 is unusable for my workloads and 12 feels like they’re just fucking with us
“640k ought to be enough for anyone.”
It’s amazing that with transparent huge pages in Linux I can have memory pages bigger than that entire 640k.
My CPU’s L1 cache is larger than that 640K system’s main memory.
EMACS: Eight Megabytes And Constantly Swapping
i could run wing commander 2, ultima underworld, and lands of lore. who needs any more?
for the same price other laptops sell for 64GB
what a bargain
…in 2030.
My sister just bought a MacBook Air for college, and I had to beg her to spend the extra money on 16gb of memory. It feels like a scam that it appears cheap with the starting at price, but nobody should actually go with those “starting at” specs.
Yeah it’s about future proofing. 8 GB might be okay for basic browsing and text editing now, but in the future that might not be the case. Also in my experience people who only want to do basic browsing and word editing, end up inevitably wanting to do more complex things and not understanding that their device is not capable of it.
Exactly. I told her that 8gb might be fine for a year or two, but if she wants this thousand plus dollar laptop to last four years she needs to invest the extra money now. Especially once she told me she might want to play Minecraft or Shadow of the Tomb Raider on it
Let me know how many multiple thousands of dollars it’s going to cost for a MAX variant of the chip that can run three external monitors like it’s 2008.
Wasn’t that only the M1 specifically that lacked that feature?
Not saying it was acceptable but pretty sure all chips after have supported 3+ monitors
My last job issued me an M2 air that could only power 1 external monitor. Was annoying as hell.
Nope. All base Mx Series Macs can only support a single external monitor in addition to their internal one.
Pro Series are professional enough that Apple deems your work worthy of using two (2) external monitors.
Max Series are the only ones that have proved their Maximum enough to Apple to let them use 3 monitors.
It’s honestly absurd. And none of them support Display Port’s alt mode so they can’t daisy chain between monitors and they max out at 3, whereas an equivalent Windows or Linux machine could do 6 over the same Thunderbolt 3 connection.
Windows and Linux machines also support sub pixel text rendering, so text looks far better on 1080p and 1440p monitors.
I have to use MacOS for work and while I’ve come to accept many parts and even like some, their external monitor support is just mind numbingly bad.
I guess you could get an eGPU. Probably not cheaper than just giving Apple their pound of flesh, though.
If apple supported egpus, sure, you could. But they don’t so…
I’m sure I remember reading about Macs using eGPUs.
https://support.apple.com/en-gb/102363
They supported it on Intel processors.
looks more
Oh, their ARM-based machines can’t.
https://discussions.apple.com/thread/254655917
No. The Apple Silicon Macs do not currently support external GPUs and Apple has not made any mention of planning to support them in the future.
sub pixel text rendering, so text looks far better on 1080p and 1440p monitors.
Why would you need that? Buy an Ultra Pro Retina Max Display and please get the stand if you don’t want Apple to go out of business.
What you’re describing as “DisplayPort alt mode” is DisplayPort Multi-Stream Transport (MST). Alt mode is the ability to pass native DisplayPort stream(s) via USB-C, which all M chip Macs are capable of. MST is indeed unsupported by M chip hardware, and it’s not supported in macOS either way - even the Intel Macs don’t support it even though the hardware is capable of it.
MST is nice for a dual WQHD setup or something (or dual UHD@60 with DisplayPort 1.4), but attempt to drive multiple (very) high resolution and refresh rate displays and you’ll be starved for bandwidth very quickly. Daisy-chaining 6 displays might technically be possible with MST, but each of them would need to be set to a fairly low resolution for today’s standards. Macs that support more than one external display can support two independent/full DisplayPort 1.4 signals per Thunderbolt port (as per the Thunderbolt 4 spec), so with a proper Thunderbolt hub you can connect two high resolution displays via one port no problem.
I agree that even base M chips should support at least 3 simultaneous displays (one internal and two external, or 3 external in clamshell mode), and they should add MST support for the convenience to be able to connect to USB-C hubs using MST with two (lower-resolution) monitors, and support proper sub-pixel font anti-aliasing on these low-DPI displays (which macOS was perfectly capable of in the past, but they removed it). Just for the convenience of being able to use any random hub you stumble across and it “just works”, not because it’s necessarily ideal.
But your comparison is blown way out of proportion. “Max” Macs support the internal display at full resolution and refresh rate (120 Hz), 3 external 6K 60Hz displays and an additional display via HDMI (4K 144 Hz on recent models). Whatever bandwidth is left per display when daisy-chaining 6 displays to a single Thunderbolt port on a Windows machine, it won’t be anywhere near enough to drive all of them at these resolutions.
Agreed, I typed quickly before bed and meant MST not alt mode.
But otherwise you’re just arguing that it’s not a big deal because ‘you don’t need any of these fancy features if you throw out your monitor every three years and buy new thousand dollar ones’.
For everyone who doesn’t want to contribute to massive piles of e-waste, we still have 1080p and 1440p, 60Hz monitors kicking around, and there is no excuse for a Mac to only be able to drive one of them with crappy looking text. It could easily drive 6 within the bandwidth of a 4k, 120Hz signal. Hell it could drive 8 or more if you drop the refresh down to 30.
I’m not generally arguing it’s not a big deal. I’m actually saying the regular M chips should be upgraded to M “Pro” levels of display support. But beyond two external displays, yes, I’m arguing it’s not a big deal, simply because >99% of users don’t want to use more than two external displays (no matter the resolution). Even if I had 6 old displays lying around I would hardly use more than two of them for a single computer. And as long as I’m not replacing all 6 displays with 6 new displays it doesn’t make a difference in terms of e-waste. On the contrary I’d use way more energy driving 6 displays simultaneously.
I’m 100% with you that MST should be supported, but not because driving six displays (per stream) is something I expect many people to do, but because existing docking solutions often use MST to provide multiple (2) DisplayPort outputs. My workplace has seats with a USB-C docking station connected to two WQHD displays via MST, and they’d all need replacing should we ever switch to MacBooks.
And sure, they should bring back proper font rendering on lower resolution displays. I personally haven’t found it to be too bad, but better would be … better, obviously. And as it already was a feature many moons ago, it’s kind of a no-brainer.
Dude I think you need to do a little research. I am typing this to you on my M1 Max (Mac studio) with 4 monitors…2 of them are at 144hz
A cursory Google search will show you that a lot of the information you threw up there is incorrect. It does seem you’re half-right about the M2 supporting only one external - the mbair does 1, but the pro does 2.
The M3 MbAir supports 2 so definitely not the case that all base Mx models support 1.
You need to reread my comment where I point out that it’s only the Max chips that can drive more than two external monitors.
And bro, a cursory Google search would also bring up this page from Apple which confirms everything I wrote. A base M3 mac can only drive two monitors if the internal display is closed, i.e. it can only drive one external monitor and one internal.
“Nope. All base Mx Series Macs can only support a single external monitor in addition to their internal one.”
Contradicts what you’re saying.
“Max Series are the only ones that have proved their Maximum enough to Apple to let them use 3 monitors”
I’m using 4.
My reading comprehension is just fine.
Read up to the part of the comment where I set the context as “in addition to their internal one”.
Re-read it or move on. I’m done. Feel free to have the last word. No desire to get into this endless quoting and telling each other what the other said nonsense.
I always thought 8gb was a fine amount for daily use if you never did anything too heavy, are apps really that ram intense now?
Yes. Just as 4GB was barely enough a decade ago.
I usually find myself either capping out the 8GN of RAM on my laptop, or getting close to it if I have Firefox, Discord and a word processor open. Especially if I have Youtube or Spotify going.
Most of that is discord, they can’t manage a single good thing right Use more GPU than the game I’m playing? Check. Have an inefficient method of streaming a game? Check. Be laggy as fuck when no longer on GPU acceleration when lemmy and guilded is fine? Check.
RAM is cheaper than my time.
I kinda consider 32GB as a minimum for anyone working on my team.I can get over 8 GB just running Discord, Steam, Shapes2
I am pretty sure most of that is just discord.
Imagine how much more room we’d have if everything wasn’t dragging a big trailer full of Chrome behind it.
I’m pretty sure Chrome doesn’t even use the memory for anything it just likes it allocated.
Imagine how far you can go on 8GB of RAM if every piece of software were still well optimized and free of bloat.
Stop. You’re scaring todays companies. Optimization? That’s a no-no word.
Now please eat whole ass libraries imported for one function, or that react + laravel site which amounts to most stock bootsrap looking blog.
Recently I downloaded Chrome for some testing that I wanted to let separate from my Firefox browser. After a while I realized my computer was always getting hot every time I opened chrome. I took a look at the system monitor: chrome was using 30% of of my CPU power to play a single YouTube video in the background. What the fuck? I ended up switching the testing environment over the libreWolf and CPU load went down to only 10%.
I’d say to try chromium, but you basically need to compile it yourself to get support for all the video codecs.
Yep. I work in IT support, almost entirely Windows but similar concepts apply.
I see people pushing 6G+ with the OS and remote desktop applications open sometimes. My current shop does almost everything by VDI/remote desktop… So that’s literally the only thing they need to load, it’s just not good.
On the remote desktop side, we recently shifted from a balanced remote desktop server, over to a “memory optimised” VM, basically has more RAM but the same or similar CPU, because we kept running out of RAM for users, even though there was plenty of CPU available… It caused problems.
Memory is continually getting more important.
When I do the math on the bandwidth requirements to run everything, the next limit I think we’re likely to hit is RAM access speed and bandwidth. We’re just dealing with so much RAM at this point that the available bandwidth from the CPU to the RAM is less than the total memory allocation for the virtual system. Eg: 256G for the VM, and the CPU runs at, say, 288GB/s…
Luckily DDR 4/5 brings improvements here, though a lot of that stuff has yet to filter into datacenters
Heavily depends on what you use, on a Linux server as a NAS I’m able to get away with 2gb, an orange pi zero 3 1gb but it essentially only ever ones one app at a time.
Im sure a hardcore rgb gamer could need 32gb pretty quick by leaving open twitch streams, discord, a couple games in the background, a couple chrome tabs open all on windows 11
And it probably won’t be able to be upgraded by the user, which should be the bare minimum.
And it
probablyabsolutely guaran-fucking-teed won’t be able to be upgraded by the user, which should be the bare minimum.You loose performance by making RAM upgradeable, hope the new RAM design, where you can install ram as if it was soldered in, is coming soon:
https://www.tomshardware.com/pc-components/motherboards/what-is-camm2
You lose performance if you can’t upgrade.
The amount of performance you lose is negligible compared to the amount of performance you lose in 5 years when the laptop’s processor is out of date and there’s barely enough ram to run a semi-heavy task.
MacBooks have an SoC, so it doesn’t make sense for the ram to be upgradable.
An SOC is a bad idea for laptops in my opinion, it just makes it harder to repair and less modular. I understand that it helps to compact the device, but it should only be used in phones if at all.
I disagree. SoCs allow for lower power consumption and a larger battery. They definitely have disadvantages but I don’t see ARM computers ever using anything but SoCs.
🆗
and maximum since you probably won’t be able to upgrade it since silicon doesn’t allow upgrades
Yeh can upgrade them at purchase. From 256gb storage to 512gb will only cost you one kidney.
It’s not an upgrade though it’s just a different model. They’re not modules you can install and I don’t even think Apple can install them you just get a different motherboard.
Which is objectionable for so many reasons, not least of all E-Waste.
Yeh I get that. Its treated as if its an upgrade - a sales upsell to a different unit I guess, rather than an upgrade to the literal unit the customer is receiving. Yep objectionable all round.
My point is you cannot effectively upgrade after the fact. You have to buy a whole new device.
Indeed. Making that initial decision even more of a forced decision toward the expensive upsell. Its evil. And wasteful as you said.
There’s reasons behind this. LPDDR IIRC works most efficiently when it’s closer to the CPU than what dimms would allow for.
Boosts speed and lowers the power requirements.
It also incentivizes people to buy larger SKUs than they originally wanted, which, bluntly, is probably the main driver for going that direction… I’m just saying that there’s technical reasons too
The technical benefits are honestly quite overblown. The M-series didn’t get the massive speed lift because it moved to soldered RAM near the CPU, it got the massive speed lift because it doesn’t have to copy stuff between the CPU and GPU, the proximity to the CPU is a pretty modest improvement. So they could’ve gotten 95% of the benefit while still offering socketed RAM, but they decided not to, probably to drive prices up.
There’s actually an argument that makes the point of driving prices down with soldered RAM.
The individual memory chips and constituent components are cheaper than they would be for the same in a DIMM. We’re talking about a very small difference, and bluntly, OEMs are going to mark it up significantly enough that the end consumer won’t see a reduction for this (but OEMs will see additional profits).
So by making it into unupgradable ewaste, they make an extra buck or two per unit, with the added benefit of our being unupgradable ewaste, so you throw it out and buy a whole new system sooner.
This harkens back to my rant on thin and light phones, where the main point is that they’re racing to the bottom. Same thing here. For thin and light mobile systems, soldered RAM still saves precious space and weight, allowing for it to be thinner and lighter (again, by a very small margin)… That’s the only market segment I kind of understand the practice. For everything else, DIMMs (or the upcoming LPCAMM2)… IMO, I’d rather sacrifice any speed benefit to have the ability to upgrade the RAM.
The one that ticks me off is the underpowered thin/lights that are basically unusable ewaste because they have the equivalent of a Celeron, and barely enough RAM to run the OS they’re designed for. Everything is soldered, and they’re cheap, so people on a tight budget are screwed into buying them. This is actually a big reason why I’m hoping that the windows-on-ARM thing takes off a bit, because those systems would be far more useful than the budget x86 chips we’ve seen, and far less expensive than anything from Intel or AMD that’s designed for mobile use. People on a tight budget can get a cheap system that’s actually not ewaste.
Has anyone ever successfully de-soldered Apple RAM and replaced it?
I think it’s proprietary ram as well so you can’t just get something off the market and solder it on. It has to be their ram or it won’t work.
dosdude on YouTube I think has done this
Pretty sure it is baked in as part of the SOC, not soldered on after the fact?
Isn’t the RAM inside the actual SoC with the Apple Silicon line? I haven’t really opened any of 'em up.
As for older Macs - sure, I know someone who replaced 8 gigs with 16 on either an Air or Pro model that had 16 available as an option but was shipped with 8. It’s just something you do when you have way too many Mac boards lying around at work and your bosses say you can’t get a new work laptop.
I don’t really care unless it has the same price point as the 8gb one.
But we all know it won’t be.
£15 more RAM. £150 more price.
$1000-$1100 is still a lot to ask of me specifically, but that is closer to market IMHO
Naturally the price for the cheapest model will also be going to up several orders of magnitude more than the cost of materials, labor, and healthy profit margin to account for that as well I’m sure.
In 1999, the iBook was US$1599 (equivalent to $2925 in 2023) (source).
The 2010 13" Air was $1299 (more in today’s $) (source).
The current 13" M3 Air is $1099 (source).
So yeah, they may well raise prices, but the cost of Apple’s entry-level hardware has decreased in absolute terms over the years, and has decreased substantially if inflation is taken into account. Not to say the margins aren’t higher (no idea about that), but it’s interesting.
Yeah it’s when you need a reasonable amount of RAM or disk that they really bend you over.
Whoa, that’s like 32GB of Windows RAM. Seems excessive to me tbh
Ironically, it’s the other way around, since Apple has to share their RAM between GPU and CPU, where other computers typically have them separately.
So in normal usage with 8 GB, you’re automatically down to 7, since at least 1GB would be taken by the graphics card. More if you’re doing anything reasonably graphics-heavy with it.
The annoying thing is I have had people claim that 8GB and 16GB is fine on Apple and works better than on PC laptops. To the point one redditor point blank refused to believe I owned an Apple laptop. I literally had to take a photograph of said laptop and show it to them before they would believe me about the RAM capacity.
I own a 8GB MacBook Pro for work, it’s definitely better than a PC with 8GB of RAM, but not better or even close to a PC with 16GB. Just the amount of stutters/freezes while the swap file goes is insane
Maybe this is true if you use Windows. If you use Linux on your PC versus macOS on a MacBook you will probably find the PC performs comparably if not better.
Oh totally, Linux is in the same ballpark as, if not better than, Macs when it comes to RAM usage. Windows is just a hog
Until you open a web browser or an Electron app. Them folks don’t really seem to give a shit about RAM usage.
A Windows application and a Mac application will use pretty much the same amount of memory regardless of operating system.
The real issue is how much memory the OS uses up. Windows is a massive waste of RAM but not enough to make any difference, certainly not with 8 GB versus 16 GB. You’re still better off on PC then.
We are talking about PC vs Mac. Both have the same problem when it comes to chromium based things.
That’s probably all the telemetrics in Windows taking their toll.
You should have said “sure buddy” and ignored them.
Obviously it depends on the situation but sometimes it is worth talking to idiots not because you have any chance of changing their mind but just demonstrate to everyone else in the thread that they are in fact an idiot. Just in case somebody thinks they have a point.
Sure buddy.
The problem is they will then keep spreading misinformation.
And that’s your problem? Do you hold $AAPL
…or $SPY, or $QQQ, or…
What even are those?
Apple helped spread that misinformation. So why would I hold some of their stock if I am trying to counter it?
No, I want companies to stop spreading this bullshit and for people to stop falling for it. I don’t hold any stocks at all. In fact that kind of bullshit I am fairly against.
AI Models needs that RAM to work
My Linux machine has 64 GiB of RAM, which is like 128 GiB of Mac RAM. It’s still not enough
Serious question what are you using all that RAM for? I am having a hard time justifying upgrading one of my laptops to 32 GiB, nevermind 64 GiB.
If games, modding uses a lot. It can go to the point of needing more than 32gb, but rarely so.
Usually, you’d want 64gb or more for things like video editing, 3d modeling, running simulations, LLMs, or virtual machines.
I use Virtual Machines and run local LLMs. LLMs need VRAM rather than CPU RAM. You shouldn’t be doing it on a laptop without a serious NPU or GPU, if at all. I don’t know if I will be using VMs heavily on this machine or not, but that would be a good reason to have more RAM. Even so 32 GiB should be enough for a few VMs running concurrently.
and run local LLMs.
Honestly, I think that for many people, if they’re using a laptop or phone, doing LLM stuff remotely makes way more sense. It’s just too power-intensive to do a lot of that on battery. That doesn’t mean not-controlling the hardware – I keep a machine with a beefy GPU connected to the network, can use it remotely. But something like Stable Diffusion normally requires only pretty limited bandwidth to use remotely.
If people really need to do a bunch of local LLM work, like they have a hefty source of power but lack connectivity, or maybe they’re running some kind of software that needs to move a lot of data back and forth to the LLM hardware, I think I might consider lugging around a small headless LLM box with a beefy GPU and a laptop, plug the LLM box into the laptop via Ethernet or whatnot, and do the LLM stuff on the headless box. Laptops are just not a fantastic form factor for heavy crunching; they’ve got limited ability to dissipate heat and tight space constraints to work with.
For me in particular I’m a software developer who works on developer tools, so I have a lot of tests running in VMs so I can test on different operating systems. I just finished running a test suite that used up over 50 gigs of RAM for a dozen VMs.
Same, 48c/96t with 192gb ram.
make -j is fun, htop triggers epilepsy.
Few vms, but tons of Lxc containers, it’s like having 1 machine that runs 20 systems in parallel and really fast.
Have containers for dev, for browsing, for wine, the dream finally made manifest.
Photo Editing, Video Transcoding.
Any memory that’s going unused by apps is going to be used by the OS for caching disk contents. That’s not as significant with SSD as with rotational drives, but it’s still providing a benefit, albeit one with diminishing returns as the size of the cache increases.
That being said, if this is a laptop and if you shut down or hibernate your laptop on a regular basis, then you’re going to be flushing the memory cache all the time, and it may buy you less.
IIRC, Apple’s default mode of operation on their laptops these days is to just have them sleep, not hibernate, so a Mac user would probably benefit from that cache.
Does it?
Previous benchmarks have shown the 8 GB models seriously fell behind in performance.
Yeah I think the joke just flew over your head.
Apple keeps saying that their RAM is somehow magic and therefore better than Windows RAM, which is a comment that obviously makes no sense.
I think they are able to share it with the GPU or something? It is maybe slightly better but it sure as fuck is not 2x better.
8 GB, even if it is “magic RAM,” is a joke amount and has been for a long time.
That’s just an APU, see consoles and laptops. The unified memory is basically just the above, but Apple also claims that due to Apple Silicon having the storage controller on board, the swap is magically faster 🤷
Also Mac OS/Linux use less RAM than Windows which certainly helps.
8GB is “fine™” on a MacBook Air, but it’s criminal for a Pro machine, and it certainly should not cost £200 for an extra 8GB. That’s genuinely insane pricing
That’s the real issue, isn’t it? The upgrade prices are disconnected from reality by a lot. If they were within the realm of sanity nobody would care much that the base is 8 GB.
I was saying this and my girlfriend when they first came out the whole thing is completely out of spec for everyone regardless of your use case.
She really only wants it for playing The Sims but you’ll run into RAM limitations there, and as you say it’s not worth paying so much more just to get a device that’s actually functional.
If you want to use it for basic word processing then you really don’t need that level of latency and you really don’t need a CPU of that level of performance. You’re just paying for stuff you’re never going to use.
If you want it for gaming there isn’t enough memory to make it worthwhile.
If you want it for intensive graphics editing work then there really isn’t enough memory for that to work.
If you want it for advanced computation then you’re probably not going for a laptop anyway. The M2 chip is obsessed with retaining battery life, which is fine in a laptop but if you want high performance applications you just want it to use more power.
It for some bizarre reason you wanted to do AI research on a laptop it’s not too bad but you’d still need the pro version and there are better things on the market but it wouldn’t be the worst I guess.
So outside of one very niche scenario it’s literally a pointless device for 99% of the user base.
In the end we got a framework laptop, which is more than capable of doing what we wanted and didn’t cost anywhere near as much. Plus it basically looks like a MacBook too. So even going to build quality wasn’t a consideration. I got one too for no particular reason, and it still ended up cheaper.
Yeah I think the joke just flew over your head.
I realize this should be a joke, but I am still unsure if it is.
It is 100% a joke. Literally other than Windows being slightly more RAM hungry, there’s not a huge difference between it and Mac’s RAM
Memory is memory. If an application requires a lot of memory then it really doesn’t matter what speed that memory is it’s more important that there’s enough of it.
There are plenty of applications that could theoretically run on the M2 MacBook in terms of processing capacity but can’t run because there isn’t enough RAM available. Oh they run in switching mode, which is super bad, because a, it’s incredibly slow, and b, it’s bad for the hard drive.
Is it like SI RAM vs US Customary RAM?
Yes. Freedom RAM equals approx. 1.6 metric RAMs. Unless your computer is on water, in which case it’s 1.857
Is that calibrated against the Universal Prototype Kilobyte in Paris?
iirc that one is outdated as it’s 1024 bytes. They haven’t been able to shave off the extra 24 bytes
1 GB is equivalent to 127,549,349,788 boron atoms.
