Apple Silicon Unified Memory for LLMs: How Much Do You Need?
How Apple Silicon unified memory fits LLMs: a tier-to-max-model table, why macOS reserves part of the pool, and how much you can really give a model.
Apple Silicon Unified Memory for LLMs: How Much Do You Need?
Apple Silicon Macs are attractive for local models because the memory is unified: the CPU, GPU, and neural engine share one pool. That removes the copy step between system RAM and VRAM that slows discrete-GPU setups. The catch is that the operating system wants part of that pool too, so the number printed on the spec sheet is not the number your model gets.
Unified memory versus discrete VRAM
On a PC with a graphics card, the GPU has its own dedicated memory and the CPU has separate RAM. A model must fit entirely in the GPU’s memory to run fast, which is why a 24 GB card caps you hard. On Apple Silicon, there is no separate GPU memory; the GPU reads from the same pool as everything else. A 70B model that would never fit one card can load as a single allocation on a 128 GB Mac.
This is the central reason the run 70B locally page lists unified-memory Macs as a valid path. The trade is peak speed: a top discrete GPU still computes faster per token than the integrated GPU in a Mac, but the Mac can simply hold bigger models.
How much you can really give a model
macOS reserves memory for the system, window server, and background processes. The exact reserve varies with the OS version and what else is open, so treat these as planning figures rather than guarantees.
| Nominal unified memory | Rough usable for model | Largest useful model (Q4_K_M) |
|---|---|---|
| 16 GB | ~11 GB | 7B, tight |
| 24 GB | ~18 GB | 13B |
| 32 GB | ~25 GB | 13B comfortable |
| 36 GB | ~28 GB | 13B, room for long context |
| 48 GB | ~39 GB | 34B, not 70B |
| 64 GB | ~54 GB | 70B, tight |
| 96 GB | ~84 GB | 70B comfortable |
| 128 GB | ~115 GB | 70B at higher quant, or 120B-class |
The “usable” column is an estimate; close other apps and you approach it, open a browser with many tabs and you fall short. Activity Monitor’s Memory tab shows the real free memory on macOS.
Why you must not give it all to the model
Two reasons keep the model below the nominal total:
- System reserve. macOS itself needs several gigabytes just to stay alive. Starve it and the machine swaps or the load is killed.
- Headroom for the KV cache. Long context grows the KV cache, as shown on the context length cost page. If you load a model that fills the pool exactly, the first long reply can push past the limit and crash.
A safe rule is to leave at least 10 to 15 percent of the pool free after the weights. On 64 GB, that means planning around a 54 GB model ceiling, which is why 70B at Q4_K_M (about 42 GB) fits but 70B at Q6_K (about 60 GB) does not.
Picking a tier
- For chat with 7B or 13B models, 24–32 GB is plenty and cheap.
- For 34B-class models at a good quant, 48 GB is the practical floor.
- For a real 70B workflow, 64 GB is the minimum and 96 GB is where it stops feeling constrained.
- For 120B-class or high-quant 70B, only 128 GB gives room.
The quantization comparison shows how much each tier saves, which is how you stretch a given Mac tier to a bigger model.
Verifying on your own machine
Do not trust the spec sheet alone. Before loading a big model, check Activity Monitor’s Memory tab or run the macOS memory command. If the “memory used” plus your planned model size is near the pool total, step down a quant tier or close apps. The unified pool is generous, but it is still a budget.
Frequently asked questions
Is Apple Silicon memory the same as VRAM?
It is unified memory, one pool shared by the CPU, GPU, and neural engine. The model and the operating system draw from the same total, which is different from a discrete GPU with its own dedicated VRAM.
Can I give the whole pool to a model?
No. macOS and background processes need part of it, so the usable amount for a model is smaller than the nominal number. On a 64 GB Mac the realistic ceiling for a model is often in the 50s of gigabytes.
Which unified memory tier runs a 70B model?
A 70B at Q4_K_M is about 42 GB with overhead, so it fits on a 64 GB machine with care and comfortably on 96 GB or 128 GB. The 48 GB tier is too tight for a useful 70B.
Why choose Apple Silicon over a discrete GPU?
The large unified pool lets you load models that would need an expensive multi-GPU PC in a single machine, at the cost of lower peak token speed than a top discrete card.
Does the GPU compete with the CPU for memory?
They share the pool, so there is no copy step between CPU RAM and GPU VRAM. That is the main advantage: big models load as one allocation instead of being split across separate memories.