Hacker News Reader

Comment by cryo32 | Hacker News Reader

I suspect that these prices are going to seriously dent sales. RAM is getting crushed. I bet the next step is going to be dumb terminals and centralisation onto all the hardware that the cloud companies bought up for AI and found wasn't possible to get any ROI out. Bezos was all over that already.

We are truly entering the dark ages of personal computing.

View on HN

Personal anecdote on ROI - I was at an early stage startup earlier this year where we had some burstable long-running GPU tasks (<100 VMs). Accross GCP and OCI we couldn't get our hands on L40S on-demand, and had to resort to T4s (released 2018). Sometimes even these were unavailable, and we would have a P4 (2016!) fallback. AWS sells A100s (2020) at $4/hr except they don't even have capacity for x1 versions, you have to rent x8.

AWS runs a hell of a lot of old junk. I was surprised at how we managed to save a lot of money not using it as well.

Supply chain crunches are not unique or new. It happens. Earth is flooded with powerful smartphones, Mac’s are already on M5 generation. Most people already get most of their computing from their phone. We will be fine.

You mean the same phones that we own less and less with each passing day? I cannot even turn off OS updates anymore. Is it even my phone if I can't do whatever I want with it?

What does this have to do with this thread? Go buy any other device then. My point is the doom and gloom is overblown, we have massively powerful devices already, no dark age is coming.

This has no bearing on your perception of ownership of your mobile device.

Pixels (and Motorolas now?) are nearly the only devices left with unlocked bootloaders.

And thanks to Google Play Integrity, even if you do liberate your device from megacorp control, you still don't get to actually use it.

"Go buy any other device" is not working out. (There should be some laws to rectify this, imo.)

The latest Google Pixels are slower than a 12 Pro iPhone and barely ahead of a 11 Pro processor wise I think Apple is safe...

You think the state of personal computing is a-ok because most people have smart phones. Others pointing out the very real limitations of what you can do on those phones seems highly relevant.

How is it not related? You are saying that we have massive powerful devices (smartphones) that can take over personal computing. I'm telling you that for you to be able to do "personal computing" you need a device that you own and can modify/run whatever you want.

The MacBook Neo went from $599 to $699. That's still significantly more powerful than anything you could buy at that price point last year.

I’m not happy with the price increases either, but saying this is the end of personal computing or that the next step is dumb terminals for everyone is very the-sky-is-falling.

Get an XPS13 for the same money, and put Linux on it. It's a much better hardware/software combo, and Apple can't unilaterally kill it by refusing to provide upgrades a few years from now.

Yeah it'll just be dead on arrival instead

Edit: Also Intel Core 5 320 that's way slower than the Neo, with same amount of RAM

> Also Intel Core 5 320 that's way slower than the Neo, with same amount of RAM

Arrant nonsense, the Core 5 comfortably beats the A18.

https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+5+320&id...

https://www.cpubenchmark.net/cpu.php?cpu=Apple+A18+Pro+%28Ma...

Seeing the other way around on Geekbench 6, for example https://browser.geekbench.com/v6/cpu/18371082 vs https://browser.geekbench.com/v6/cpu/16884902 and the Intel has more samples there than on Passmark

n.b. the Geekbench multicore test is more like a 2-core test because it adds lock contention, which is arguably closer to the real use case for these machines but isn't what you normally think of a multicore test doing

Though it does seem safe to say your original claim of "way slower than the Neo" isn't correct. Considering it's losing in one benchmark and only ahead in Geekbench (that tends to show higher scores for Apple processors relative to other benchmarks anyway).

"Roughly equal" seems to be a more accurate description.

It's not normal for these two benchmarks to deviate that much, so I'm not gonna take an average between them, just trust one or the other. My default has always been Geekbench, I didn't even look at Passmark until it was mentioned, and in this case Geekbench thinks the Mac's single-core speed is 35% higher.

I'm hoping for a renewed focus on writing software that runs well on constrained machines and helps people get many more years of useful life out of their existing computers.

I'd love that. I'd also love to see the Jets win the super bowl. Pretty sure neither happens in my lifetime.

> I bet the next step is going to be dumb terminals and centralisation

This is one of my biggest fears of this whole thing, that personal (local) computing is going to effectively die.

I mean Micron exited the consumer market entirely. All fab capacity is going to HMB, not consumer chips. The cartel has zero desire to make consumer hardware anymore, AI/data centers are far too profitable for them. Micron just reported gross margin of 85%.

So the cartel is raking in the dough selling shovels, screwing consumers, with long term supply deals already in place, they have no need to even think about the personal computer market (or chips for anything else either, this is going to cascade into automotive and elsewhere) until at least 2028-2029.

I'm sure Microsoft is frothing at the mouth to sell people thin clients with a Windows 365 subscription, and I wouldn't be surprised to see the new XBox go all in on cloud gaming like GeForce now.

We're stuck in this situation until/unless the AI buildout slows or stops in some sort of market correction.

I know some organisations were already moving to thin clients last year. Citing cutting costs and improving security (the data doesn't stay in employee's laptop and all access to virtual desktop is thoroughly centrally logged).

Massive pushback (lagging, accessibility issues, slow) from workers was ignored and many people quit.

We aren't quite there yet where I work but those conversations are starting. We've already pushed refresh cycles out for the non-tech folks from 3 years out to 5 years with justification (basically has to be broken or battery shot, otherwise its run it til it dies or no longer gets updates, no more automatic refresh).

Sucks, but can't say I disagree with the fresh times though. There hasn't been a compelling need to upgrade all knowledge workers every 3 years anyway. An M2 air from 2022 is still fine today and will likely continue to be fine for at least another 3 years or more.

A 2022 Dell laptop with windows 11 isn’t doing great in 2026 though

It's been like this for over a decade, and for legit reasons.

I'm sure they've done the math. Mac has ~8% revenue share for Apple and I (naively) assume they'll just account for a 20% drop in sales with 20% higher prices. Personally, if my Mac were to die right now, I'd scream and shout (well, I'd use Apple Care...), but I won't go back to a Linux laptop, since I'm too deep into the ecosystem. And I suspect I'm not alone.

fwiw, I don't hate the thin-client model for dev work (via ssh, certainly not RDP - I've done both), but I despise the implications of _having_ to do it.

Same boat here. If I had to, I could grab an Air instead and do more work over ssh. I prefer to keep things local, but it's not a huge deal breaker for my work. I'm too deep in the ecosystem to get anything else, and I need Xcode anyway.

I suspect a lot of mac users are in the same place, and Apple knows it.

The current analyst estimate is a 15% decline in sales.

GPU farms aren't that useful for general purpose work

Yeah but there's a huge amount of generic estate to support those GPUs.

Reply on news.ycombinator.com