frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Ask HN: MacBook vs. Dedicated GPU for LLM

15•mzubairtahir•1h ago
For those who are using llms on macbook, Want to understand how macbook is different than dedicated GPU in running those models? and how to know how much a macbook is capable of running a model?

Comments

JSR_FDED•1h ago
MacBooks with their unified memory behave like a slow GPU with enormous amount of video RAM. So you can run large smart models slowly.

Dedicated GPUs have less video RAM so can run smaller less smart models quickly.

exabrial•59m ago
Do Mac Pros provide more headroom? noob here, noob questions
rho138•40m ago
Idk why you’re being downvoted for asking a question. Pending specs they _could_ provide more headroom for a larger model but they would still be limited by the CPU and it’s associated bus speeds.
JSR_FDED•5m ago
In what sense? Headroom for what?
mzubairtahir•33m ago
how much memory is actually useable by gpu in macbook? as it is shared?
pylotlight•18m ago
roughly ~50–56GB, although this is somewhat configurable with iogpu.wired_limit_mb. By default, macOS reserves ~25% of memory for the system.
visarga•29m ago
Macbook M5 64GB - can run gemma-4-26b-a4b-it-4bit and Qwen3.6-35B-A3B-4bit at about 1500 tps prefix and 45 tps decode on contexts up to 100K tokens using MLX. It's faster than Claude. I was really surprised, chat quality is also similar to Claude for gemma4. Agentic works but does not compare to cloud models, you can still make agents where top level is code.
mzubairtahir•27m ago
sorry but asking again: how much memory is actually useable by gpu in macbook? as it is shared(os and apps also have to use same memory)? and it is different than dedicated gpu memory?
gizajob•16m ago
It’s completely shared so the OS and everything else takes up maybe 8GB of the RAM. On a 64GB machine you can run models about 45GB in size and still have space for those models run other tasks which themselves might need ram. To a user, the GPU appears to just use the RAM as much as it needs same as any other process running on the system. You can see what space your LLMs are taking up in Activity Monitor (or htop) and how much GPU capacity they’re using (all of it)
epsteingpt•1h ago
Both are going to be super super slow and low payback.

You gotta really want it right now.

It's still early!

jpgvm•43m ago
If you want a massive MacBook anyway then it's great. They are decent for local LLMs, awesome for local image models and it's a MacBook so AppleCare+ has your back. IMO it's a no brainer if you wanted a MacBook anyway but it's a poor choice if your reason to buy it is to run LLMs.
mzubairtahir•31m ago
are you saying because of speed or it just cant run them?
zepearl•2m ago
[delayed]
brcmthrowaway•42m ago
Dual 3090 >>> Any Apple product.
gnabgib•36m ago
> Dual 3090 >>> Any Apple product.

Dual 3090s are terrible airpods

tim-tday•15m ago
I snorted
kylec•33m ago
Doesn't the 3090 cap out at 24GB VRAM? That's not a lot to run a local model
mzubairtahir•29m ago
but still it can run handsome models
nichch•41m ago
My opinion is that you should wait for 6-12 months before making a purchase either way.

Open weight models are getting good. With GLM 5.2 now chasing Opus, I'm very excited to see a smaller model's distillation.

Plus, the OLED MacBook Pro should be released by then.

Frannky•14m ago
This is my opinion too. Even if you buy hardware like a cluster of 8xGB10s or 4 A100s, they'll still be slow and a little dumber than what you're used to. We need to wait a little for better hardware. Lots of companies are pushing the frontier, so hopefully it'll come very soon.

Competition and innovation will hopefully make the bubble pop, and we'll get reasonably priced local hardware to run very intelligent models. Something like Talaas with GLM 5.2 would be pretty cool. Or Apple printing the latest model onto hardware—it would give a new reason to buy a new Mac every year (a new ai model with every new version).

gizajob•4m ago
The hardware is here today for people prepared to tolerate mild amounts of latency. It’s easy to forget that computing tasks used to often take major amounts of time - rendering an audio file, rendering a video, transcoding – all kinds of tasks took minutes or even hours of the computer spinning its fans on maximum just to deliver the result. AI and agentic AI and diffusion is the next round of that - trading a small bit of your waiting time for phenomenal power. The datacentre builders trying to get you hooked on instant responses on the LLM platforms have made you think that a “good” AI responds instantly and completely interactively - they can still be brilliant with a bit of delay. And having a competent agent doing things on my local machine, it doesn’t really matter if it takes ten minutes or an hour or six hours to complete a task while I’m out doing other things.
cylentwolf•37m ago
I asked a few of my friends that are ML engineers this question and all of them said to run the LLMs in the cloud with their infrastructure because it was going to be way faster. If you just want to tinker around I would look at @JSR_FDD's comment.
gizajob•27m ago
Local LLMs running in LM Studio on a MacBook Pro work great, if you’re prepared to wait for the answers because using an LLM locally is much much slower than having the instant results appear when using an online LLM like ChatGPT or Claude. You can also run OpenClaw on the MacBook and have that act as the front end for the LLM, to get full interactivity and have it install command line tools on your Mac to perform whatever tasks you’ve set it.

If you don’t already have a MacBook, then there’s a bit of a sweet-spot for the AI experimenter right now, which is to buy a second-hand 16” MBP with an M1 Max chip and 64GB of shared ram. Because these are about 5 years old now, they have depreciated to the point where they can be had for around £1100 / €1300 / $1500 and make a phenomenal platform for learning because the 64Gb of shared memory means you can host models up to about 48GB in size, and then task them to do interesting things with coding without ever having to worry about token burn.

The downside is that they’re slow, and prone to having to be nudged to keep them on track, but that’s part of the fun too. The “latency” is atrocious granted - you ask something and the machine thinks for a few minutes before saying anything which is a different experience to using Claude. But… it does work. You can think of yourself more like a manager with a junior member of staff and set the machine running and leave it to do its thing for a couple of hours which can be actually useful work, but this approach will likely be shouted down by some commenters here who treat Claude like some kind of expensive and quick-fire dopamine pump. Can also use a Mac like this for running diffusion models for image generation and suchlike in ComfyUI, even though, again, results will be slow. Spending more money on a more recent MBP with as much RAM as you can afford will deliver the same results more expensively in a quicker and quicker time.

To get the same kind of size of model you’d have to combine a couple of Nvidia 3090 24GB cards in a decent workstation with the PCI capacity to handle them, or hack some kind of solution to hang GPUs off the back of a motherboard on ribbon cables with the GPUs running on their own PSU, which is what I’m building next… the difference is those cards have 24GB of vram and cost about $1000 each second-hand, but will operate much much faster than the M1 Max MBP, or even the most recent M5 because they have so much more bandwidth (because they’re burning 350 watts on GPU compute rather than 140 watts total which is what a super efficient MBP has for the cpu/gpu/screen/everything).

So say you had $6000 to spend today, you could buy a second hand workstation and craft a solution with external GPUs which would completely smoke any Mac in existence, even though macs have the edge in the size of model you’d can run (slowly) due to their shared memory. External GPUs and access to the Nvidia frameworks and general CUDA ecosystem wins out on the performance front though. A real sweet spot is to buy an M1 Max MBP and have that as your front end to a Linux workstation full of GPUs.

But any apple silicon MBP is a totally competent gateway drug to local agentic computing.

Google Gemini could give you an in-depth and useful discussion about this exact question.

AI glasses help students cheat in exams – test-obsessed Asia is ground zero

https://www.rnz.co.nz/news/world/628396/ai-glasses-are-helping-students-cheat-in-exams-test-obses...
3•billybuckwheat•4m ago•0 comments

Certainty Volatility Theory: Stable Uncertainty vs. Volatile Certainty

https://theguptalog.blogspot.com/2026/06/adaptive-certainty-theory-maybe-we-dont.html
1•GeorgeWoff25•8m ago•0 comments

Everything* – An interactive voyage through all orders of magnitude

https://www.everything.one/#overview
1•bj-rn•10m ago•0 comments

Lore: Next-generation open source version control

https://github.com/EpicGames/lore
1•taubek•17m ago•0 comments

Gambling disorder cases skyrockets in states where sports betting is legal

https://twitter.com/cremieuxrecueil/status/2070651565107446116
1•MrBuddyCasino•17m ago•0 comments

The New York Times Amends Lawsuit Against OpenAI and Microsoft

https://www.nytimes.com/2026/06/25/technology/times-lawsuit-openai-microsoft.html
1•1vuio0pswjnm7•19m ago•0 comments

Promptetheus – Trace, detect, and auto-repair AI agent failures

https://github.com/obro79/promptetheus
1•tar-ive•20m ago•0 comments

People Inc. CEO accuses Google of abusing its market power

https://www.axios.com/2026/06/23/people-inc-google-ai-search-crawler
3•1vuio0pswjnm7•22m ago•1 comments

Ask HN: How much memory is useable by GPU in MacBook?

1•mzubairtahir•22m ago•1 comments

Rich and famous families teach their kids Mandarin as China's influence grows

https://www.abc.net.au/news/2026-06-27/musk-trump-bezos-zuckerberg-children-learning-mandarin/106...
2•defrost•23m ago•0 comments

My gf's birthday is my yearly reminder that I'm bad at gift-giving

https://giftmaxxing.vercel.app/
1•tar-ive•23m ago•0 comments

I HATED having to stop playin instrument togo thru pages onPDFso I used blinking

https://verceleye.vercel.app/
1•peenut•24m ago•1 comments

OpenTTD 16.0-Beta1

https://www.openttd.org/news/2026/06/25/openttd-16-0-beta1
1•untilted•26m ago•0 comments

Mythos available for select organisations

https://twitter.com/anthropicai/status/2070665903440871779
1•dsr12•28m ago•1 comments

Air conditioning torn from homes under net zero clampdown

https://www.telegraph.co.uk/money/consumer-affairs/air-conditioning-torn-from-homes-under-net-zer...
2•razor-thin•32m ago•1 comments

Photos by Om

https://www.photosbyom.com
1•adityaathalye•32m ago•1 comments

Modern GPU Programming for MLSys

https://mlc.ai/modern-gpu-programming-for-mlsys/#
1•sonabinu•46m ago•1 comments

Coding agent harness written in native Golang with built-in file and Git viewer

https://code.intellios.ai
1•coolwulf•46m ago•0 comments

2026 FIFA Worldcup Predictor

https://nutmeg-4eg.pages.dev/
1•chriszng•53m ago•2 comments

Isn't US Government trying to monopolize AI as a super power?

2•StizzurpXDD•59m ago•0 comments

SpaceX stock is a terrible buy – what that means for the bull market

https://www.msn.com/en-us/money/topstocks/spacex-stock-is-a-terrible-buy-what-that-actually-means...
2•WheelsAtLarge•1h ago•1 comments

Apple seeks to buy memory chips from blacklisted Chinese company

https://www.ft.com/content/d72a25e2-7bde-4aa9-bd8d-0c4f3d6cb2cb
1•ilreb•1h ago•0 comments

Agentic Trading on Robinhood

https://robinhood.com/us/en/agentic-trading/
4•huragok•1h ago•2 comments

(Planet Money, NPR) "We almost had a smartphone in the 90s. Why did it fail?"

https://www.npr.org/2026/06/26/nx-s1-5872785/general-magic-sony-magiclink-constraints
1•wnc3141•1h ago•2 comments

I built a marketplace where the money can be verified by anyone

https://old.reddit.com/r/SubliminalsVerytasium/comments/1ufhleb/i_built_a_marketplace_where_the_m...
1•bbenevolence•1h ago•0 comments

Find the right AI agents to build

https://www.agentideahub.com
1•mattmerrick•1h ago•0 comments

TUI email client in native Golang with LLM based drafting functions

https://mail.intellios.ai
1•coolwulf•1h ago•0 comments

Nomad: Portable, offline media server powered by the ESP32-S3 in a thumbdrive

https://www.instructables.com/Jcorp-Nomad-Mini-WIFI-Media-Server/
1•thunderbong•1h ago•0 comments

US allows Anthropic to release Mythos AI to 'trusted' US organizations

https://www.reuters.com/technology/us-releases-anthropic-model-mythos-some-us-companies-semafor-r...
1•swolpers•1h ago•0 comments

If they start to gatekeep who gets to use the best models, that is a DoW

https://twitter.com/jmrphy/status/2070528497752166454
1•Jimmc414•1h ago•0 comments