Apple192 GB VRAM800 GB/s$0/hr

Qwen2 72B on Mac M2 Ultra (192GB): Does It Fit, How Fast, What It Costs

Qwen2 72B at FP8 (E4M3) needs 72.7 GB for its weights and 94.2 GB in total at 8K context with 8 concurrent users. That means it fits on a single Mac M2 Ultra (192GB). Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
94.2 GB needed / 192 GB available
Decode speed
10.73
tok/s per stream
Aggregate
92.85
tok/s @ 8 concurrent
Cost / 1M tokens
—
1× Mac M2 Ultra (192GB) · quote required
What changes as you add Mac M2 Ultra (192GB)s
Tensor-parallel degree vs VRAM, throughput, and cost. Cost scales linearly with GPU count; decode throughput scales near-linearly too, so cost per token stays roughly flat.
GPUsTotal VRAMVRAM neededDecode tok/sAggregate tok/sTTFTCost / 1M tokFits?
1× Mac192 GB94.2 GB10.7392.8588.35 s—✓
2× Mac384 GB94.2 GB18.24371.444.20 s—✓
4× Mac768 GB94.2 GB36.48148622.12 s—✓
8× Mac1536 GB94.2 GB72.96594211.07 s—✓
Concurrency vs context length
On 1× Mac M2 Ultra (192GB) at FP8 (E4M3). KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K881.34 GB✓ fits
8K442.68 GB✓ fits
33K1110.74 GB✓ fits
66K521.47 GB✓ fits
131K242.95 GB✓ fits
Other GPUs that run Qwen2 72B
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB4×3922$0.1416
A100 80GB SXM42×1094$0.6092
H100 SXM5 80GB2×1833$0.7577
AMD MI300X 192GB1×853.0$0.8142
L40S 48GB2×472.8$0.8226
AMD MI325X 288GB1×965.6$1.01
Frequently asked questions

Can Qwen2 72B run on Mac M2 Ultra (192GB)?

Yes. Qwen2 72B at FP8 (E4M3) needs 94.2 GB including the KV cache, and Mac M2 Ultra (192GB) has 192 GB — so it fits on a single GPU.

How fast is Qwen2 72B on Mac M2 Ultra (192GB)?

10.73 tokens/sec per stream and 92.85 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 88.35 s, and inter-token latency is 93.2 ms.

What does Qwen2 72B on Mac M2 Ultra (192GB) cost per million tokens?

— per 1M output tokens at 1× Mac M2 Ultra (192GB). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Qwen2 72B need on Mac M2 Ultra (192GB)?

72.7 GB for the weights at FP8 (E4M3), plus 21.5 GB of KV cache for 8 concurrent requests at 8K context — 94.2 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is Mac M2 Ultra (192GB) the right choice for Qwen2 72B?

It works. Compare — per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 8× RTX 4080 16GB at $0.1062, and the fastest is 4× Cerebras CS-3 (WSE-3) at 43774 tok/s.

Adjust the workload yourself

Same formulas, your context length, concurrency, and quantisation.

Open in calculator