Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Apple caught off guard by AI demand for Mac Mini and Mac Studio

macrumors.com

Post reply on HN
Print view View original

631–640 of 640 posts

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#631

Post by aurareturn »

adolph wrote: post #630
aurareturn wrote: post #591

Earlier quoted context omitted.

They didn't make their GPUs for crypto. Crypto companies just bought a lot of their GPUs. If they had a choice, they would have sold every GPU to gamers instead because a crypto boom was always followed by a crash which flooded the market with used GPUs and crashed Nvidia's stock price.

> They didn't make their GPUs for crypto. Sure they did and still do. They did a good job of managing the bubble but participate in the market nontheless. Read the above article: Nvidia has promised that its new lineup of cryptocurrency mining GPUs, called CMP for short, won't impact the supply of GeForce graphics cards for PC gamers. . . Since these cards are destined for mining rigs they will also lack video output…

They made a dedicated chip because miners were buying up all the GPUs from gamers.

They also nerfed GPU hash rates through software to prevent miners from buying them.

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#632

Post by aurareturn »

gwerbin wrote: post #607
f055 wrote: post #563

Earlier quoted context omitted.

I don’t get it. You don’t need a Mac mini to use Claude Code on the go, any home server with stable internet is enough. Was it the marketing?

Either they're running Claude Code with a local LLM (technically possible via dev tools albeit unsupported), or they genuinely have no idea why the Mac is good for AI and just bought it because they were told it was good for AI. The former honestly sounds like a nightmare because the Mac Mini can run some big models but not that big, and Claude Code will guaranteed stuff your context with 35k tokens, and all the Anth…

  Either they're running Claude Code with a local LLM (technically possible via dev tools albeit unsupported), or they genuinely have no idea why the Mac is good for AI and just bought it because they were told it was good for AI.
Neither. I use Claude Code subscription through my Mac Mini when I’m on the go.

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#633

Post by aurareturn »

nullbio wrote: post #617
aurareturn wrote: post #543

Earlier quoted context omitted.

I'm convinced that this is just guerilla marketing from Apple. I'm not. I have an M4 Mac Mini running 24/7 so I can use it as my Claude Code machine when I'm on the go. Why would they do a guerilla marketing campaign now if their Macs are generally sold out despite nearly doubling the Mac Mini price in the last few months?

So your argument against this being guerilla marketing is that you bought a Mac Mini after seeing the guerilla marketing, to do the thing that you didn't need a Mac Mini to do.

No I bought it over a year before this “marketing”.

I’ve found a new use for it which I’m assuming others have as well.

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#634

Post by kristianp »

Dylan16807 wrote: post #469
Grombobulous wrote: post #394

Earlier quoted context omitted.

I also think it’s possible that it may be best to continue waiting on the Mac side of the house despite the increasing prices. I think that some of the hardware design folks have been blindsided by AI demand and we haven’t really gotten that next generation AI hardware yet, to the point where buying M5 isn’t going to make sense in a couple of years. Rumors seem to be that the M7 is the generation that Apple is lookin…

> Rumors seem to be that the M7 is the generation that Apple is looking to push AI performance much further. What specifically does that mean? The M5 series has 10 cores per 128 bits of memory bus; are the cores unable to keep up with the RAM? I thought they did and memory bandwidth was usually the bottleneck. But the memory bus is already very highly clocked and goes up to 1024 bits wide so it's hard to picture memo…

LLMs token generation is memory bandwidth constrained. If the M7 has double the memory bandwidth as some speculate [1], then it will help with LLM performance. There is also expectation (possibly unfounded) that the GPU will have improved matrix performance to help prompt processing as well.

[1] "LPDDR6 is coming." - https://news.ycombinator.com/item?id=49436849

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#635

Post by Dylan16807 »

kristianp wrote: post #634
Dylan16807 wrote: post #469

Earlier quoted context omitted.

> Rumors seem to be that the M7 is the generation that Apple is looking to push AI performance much further. What specifically does that mean? The M5 series has 10 cores per 128 bits of memory bus; are the cores unable to keep up with the RAM? I thought they did and memory bandwidth was usually the bottleneck. But the memory bus is already very highly clocked and goes up to 1024 bits wide so it's hard to picture memo…

LLMs token generation is memory bandwidth constrained. If the M7 has double the memory bandwidth as some speculate [1], then it will help with LLM performance. There is also expectation (possibly unfounded) that the GPU will have improved matrix performance to help prompt processing as well. [1] "LPDDR6 is coming." - https://news.ycombinator.com/item?id=49436849

14GT/s is a goal but it might be a while, I think 11-12 is more likely in that time frame.

50% more pins... we'll see. If they can reasonably make that fit then even more shame upon the traditional desktop CPU makers for sticking with 128 bits for so long.

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#636

Post by srslack »

ActorNightly wrote: post #628
srslack wrote: post #370

Earlier quoted context omitted.

I am using Mac studios for local inference, but not for openclaw. It was simply the best bang for the buck at the time, and the software side of things has gotten even better since then. I haven’t run the pricing comparison but I have a feeling with the direction they’re heading that it’s better to hold off for m6 as it may assuage various bottlenecks if you are latency sensitive.

>It was simply the best bang for the buck at the time, I wouldn't have a problem with Apple fanboys if they just stated that they like Macs, but for some reason you have no problem with just straight up LYING. Or at least being so deluded that you state things that are so easy to prove false as fact. I.e, 600 gb/sec is dogshit slow compared to vram speeds.

You could not get 500GB of VRAM for less than $100k at the time. It was the absolute best bang for the buck as far as local inference and fine tuning. I am not sure why you are so hostile when I specifically called out “latency sensitivity” as far as prompt preprocessing and various other bottlenecks. I run Linux.

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#637

Post by manmal »

ustad wrote: post #480
manmal wrote: post #427

Earlier quoted context omitted.

The Amphetamine app from the App Store keeps a MacBook awake even with the lid closed.

You need an app for that?

I guess `caffeinate` doesn't go that far because it can easily lead to overheating.

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#638

Post by ActorNightly »

srslack wrote: post #636
ActorNightly wrote: post #628

Earlier quoted context omitted.

>It was simply the best bang for the buck at the time, I wouldn't have a problem with Apple fanboys if they just stated that they like Macs, but for some reason you have no problem with just straight up LYING. Or at least being so deluded that you state things that are so easy to prove false as fact. I.e, 600 gb/sec is dogshit slow compared to vram speeds.

You could not get 500GB of VRAM for less than $100k at the time. It was the absolute best bang for the buck as far as local inference and fine tuning. I am not sure why you are so hostile when I specifically called out “latency sensitivity” as far as prompt preprocessing and various other bottlenecks. I run Linux.

The reason Im so hostile is because it personally irks me when a technical forum like HN is filled with Apple-slop being passed off as technical discussion. Its fine to like Apple for what it is, the ecosystem, the industrial design, the advantages of dedicated hardware with specialized software to extend battery life. But instead, there is Apple-slop, which is basically starting with the assumption that whatever you can do on Mac is the standard, and that Macs are great because they fit that standard.

Being able to run frontier models at <10 tok/sec is just an excersize in showing that you can afford a Mac. Its useless for any real tasks. Gemma4 is on par with a lot of the older frontier models like Gemini Flash, and you can run that at 100+ tok/sec on a 2x3090 rig which costs way less than even top of the line Mac Studio.

Re: Apple caught off guard by AI demand for Mac Mini and Mac Studio

#640

Post by qeternity »

redox99 wrote: post #621
qeternity wrote: post #284

Earlier quoted context omitted.

> Edit: I'm thinking of a headless Mac mini, if you meant running it on the same machine you're using of course you'll need more memory, but LLMs are best served from a headless server so that's what I'd recommend. What? LLMs are best served from a massive PD disaggregated cluster of B300s connected via NVLink. If you're running LLMs on a Mac Mini, it's because you want to run local, not because it's the best setup.

>massive PD disaggregated cluster of B300s connected via NVLink. So a headless server. Macs were mentioned because that's what the post is about. It could be a PC (I use a 2x3090 PC). The point is that it's a better experience to have a box dedicated to the LLM than running it in your system. Obviously in your home, so local.

Has nothing to do with being headless. That's just a natural outcome.

If someone wanted to use an 8x B300 as their daily driver...go ahead.

It would still be the best way to serve a given model.

Post reply on HN