32GB is the maximum memory configuration for the 14-inch laptop, which isn’t sufficient for running local LLMs. I think a Mac Studio or Mac Mini with higher memory would be more useful.
Apple M5 chip
221–230 of 1001 posts
Re: Apple M5 chip
#222Earlier quoted context omitted.
This is not the first time this has happened in Apple’s history. The transition from the 68k architecture to the PowerPC brought major performance improvements, but Apple’s software didn’t take full advantage of it. If I remember correctly, even after the PowerPC switch, core elements of the classic Mac OS still ran in emulation as late as Mac OS 9. Additionally, the classic Mac OS lacked protected memory and preempt…
Sixteen years ago, John Gruber wrote: > Hardware and software both matter, and Apple’s history shows that there’s a good argument to be made for developing integrated hardware and software. But if you asked me which matters more, I wouldn’t hesitate to say software. All things considered I’d much prefer a PC running Mac OS X to a Mac running Windows. https://daringfireball.net/2009/11/the_os_opportunity At the time I…
Windows would have beat MacOS only if Microsoft had just done one small, teeny-weeny thing - just left the OS alone after Win 10.
Re: Apple M5 chip
#223The big win would be a linux capable device. I don't have any interest in mac os x but the apple m parts always seem amazing. In theory this would be where qualcomm would come in and provide something but in practice they seem to be stuck in qualcomm land where only lawyers matter and actual users and developers can get stuffed.
Yeah, this is the biggest hole in ARM offerings.
The only well supported devices are either phones or servers with very little in between.
Even common consumer devices like wifi routers will have ARM SOCs with pinned version of the kernel they are attached to which will get supported for 1 to 2 years at most.
Re: Apple M5 chip
#224Re: Apple M5 chip
#225The M5 MacBook Pro still gets the Broadcom WiFi chip but the M5 iPad Pros get the N1 and C1X (Sweet). All in all, apple is doing some incredible things with hardware. Software teams at apple really need to get their act together. The M1 itself is so powerful that nobody really needs to upgrade that for most things most people do on their computers. Tahoe however makes my M1 Air feel sluggish doing the exact same task…
Yeah I love my M1 iPad Pro. But the "liquid glass" update made it feel slower. Really only the 'unlock' feels slower, once I'm using it it's fine. But it's slightly annoying and does make me want to update this year to the m5. But it's a glorified Kindle and YouTube box, so I'm hesitating a little bit.
my dad's got a pre AS iPad Pro and it's so bad after updating to 26. My 6th gen iPad on iOS 17 felt faster than this
Re: Apple M5 chip
#226And here I am, selling my Macbook M4 Pro to buy a Macbook Air and a dedicated gaming machine. I've tried gaming on the Macbook with Heroic, GPTK, Whiskey, RPCS3 emu and some native. When a game runs, the performance is stunning for a Laptop - but there is always glitches, bugs and annoyances that take out the joy. Needles to mention lack of support from any sort of online multiplayer, due to the lack of anticheat sup…
Going back to the Air's screen from your Pro will be a steep fall.
Re: Apple M5 chip
#227A unified memory bandwidth of 1,224 gigabits per second is quite impressive.
Nvidia DGX Spark has 273 GB/s (2184 gigabits with your units) and people are saying it's a disappointment because that's not enough for good AI performance with large models. All the neural accelerators in the world won't make it competitive in speed with discrete GPUs that all have way more bandwidth.
> All the neural accelerators in the world won't make it competitive in speed with discrete GPUs that all have way more bandwidth.
That’s true for the on-GPU memory but I think there is some subtlety here. MoE models have slimmed the difference considerably in my opinion, because not all experts might fit into the GPU memory, but with a fast enough bus you can stream them into place when necessary.
But the key difference is the type of memory. While NVIDIA (Gaming) GPUs ship with HBM memory ship for a while now, the DGX Spark and the M4 use LPDDR5X which is the main source for their memory bottleneck. And unified memory chips with HBM memory are definitely possible (GH200, GB200), they are just less power efficient on low/idle load.
NVIDIA Grace sidestep: They actually use both HBM3e (GPU) and LPDDR5X (CPU) for that reason (load characteristics).
The moat of the memory makers is just so underrated…
Re: Apple M5 chip
#228The M5 MacBook Pro still gets the Broadcom WiFi chip but the M5 iPad Pros get the N1 and C1X (Sweet). All in all, apple is doing some incredible things with hardware. Software teams at apple really need to get their act together. The M1 itself is so powerful that nobody really needs to upgrade that for most things most people do on their computers. Tahoe however makes my M1 Air feel sluggish doing the exact same task…
“nobody really needs to upgrade that for most things”
Maybe, but for lots of scenarios even M5 could still benefit from being an order of magnitude faster.
AI, dev, some content scenarios, etc…
Re: Apple M5 chip
#229> Apple 2030 is the company’s ambitious plan to be carbon neutral across its entire footprint by the end of this decade by reducing product emissions from their three biggest sources: materials, electricity, and transportation. But never, ever, through not shipping incremental hardware bumps every year regardless of whether there's anything really worth shipping.
I'm always skeptical about these carbon neutral pledges because in practice it's a lot of administrative magic, like paying a company that says they will plant trees or whatever which will sign some official looking paper saying 'ye apple totaly compensated three morbillion tonnes of carbon emissions'.
And it's things like not including a charger, cable, headphones anymore to reduce package size, which sure, will save a little on emissions but it's moot because people will still need those things.
Re: Apple M5 chip
#230It feels like apple is “ a square peg in a round hole” when it comes to AI - atleast for now. They are not the hardware provider like nvidia, they don’t do the software and services like OpenAI or even Microsoft/oracle. So they are struggling to find a foothold here. I am sure they are working on a lot of things but the only way to showcase them is through their phone which ironically enough feels like not the best p…
Being late in AI race or not entering it from training side is not necessarily bad, others have burned tons of money, if Apple enters with their hardware first (only?) it may disrupt status quo from consumer side. It's not impossible that they'll produce hardware everybody will want to run local models that will be on par with closed ones. If this happens it may change real money flow (as opposed to investor based on imaginary evaluation money that can evaporate).