MacBook Pro M4 Pro 64GBBudget pick
64 GB Unified (+40)
A
Adds memory headroom for longer context windows and future model growth.20.2 tok/s decode
Adds memory headroom for longer context windows and future model growth.
~$1,599 MSRP
Nous Dolphin 13B needs ~25.2 GB VRAM. Intel Arc Pro B60 24GB has 24.0 GB. With Q5_K_M quantization, expect ~19 tok/s.
Operating mode
Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
Select quantization to explore
1.2 GB over capacity — needs offload or smaller quantization
Fit status
Runs with offload (needs ~0.4 GB host RAM)
Decode
18.6 tok/s
TTFT
10394 ms
Safe context
14K
Memory
25.2 GB / 24.0 GB
The raw memory story may look fine, but the software ecosystem is still a constraint here.
Very little memory headroom
You can run the model, but there is not much room left for longer context, bigger batches, extra apps, or future model updates.
Runtime ecosystem is narrower than CUDA
Intel GPUs can look attractive on memory per dollar, but local AI tooling, kernels, and model coverage are still broader and easier on CUDA today.
Prefer CUDA if you want the path of least resistance
If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
Buy headroom, not only minimum fit
A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | A | Runs well | 26.8 tok/s | 3935 ms | 14K |
| Coding | B | Runs with offload (needs ~0.4 GB host RAM) | 18.6 tok/s | 10394 ms | 14K |
| Agentic Coding | F | Too heavy | 8.3 tok/s | 33874 ms | 14K |
| Reasoning | B | Runs with offload (needs ~0.4 GB host RAM) | 18.6 tok/s | 12284 ms | 14K |
| RAG | F | Too heavy | 8.3 tok/s | 42343 ms | 14K |
Inference speed
Estimated decode speed (tokens/sec) for Nous Dolphin 13B at Q5_K_M across popular GPUs and Apple Silicon, using the fastest local runtime per device. Fastest is RTX 5090 32GB at ~131 tok/s. Speed is memory-bandwidth bound, so cards that fit the whole model in VRAM run far faster than ones that offload to system RAM.
| GPU / Mac | Memory | Quant | Speed (tok/s) | Fits? |
|---|---|---|---|---|
 RTX 5090 32GB | 32 GB | Q5_K_M | 130.8 | Fits |
Mac Studio M3 Ultra 256GB | 256 GB | Q5_K_M | 60.7 | Fits |
| RTX 4090 24GB | 24 GB | Q5_K_M | 56.7 | Offloads |
RX 7900 XTX 24GB | 24 GB | Q5_K_M | 51.1 | Offloads |
Mac Studio M2 Ultra 128GB | 128 GB | Q5_K_M | 50.6 | Fits |
| RTX 3090 24GB | 24 GB | Q5_K_M | 48.5 | Offloads |
Mac Studio M1 Ultra 128GB | 128 GB | Q5_K_M | 47.9 | Fits |
MacBook Pro M4 Max 128GB | 128 GB | Q5_K_M | 33.0 | Fits |
MacBook Pro M4 Max 64GB | 64 GB | Q5_K_M | 33.0 | Fits |
MacBook Pro M3 Max 64GB | 64 GB | Q5_K_M | 26.2 | Fits |
MacBook Pro M1 Max 64GB | 64 GB | Q5_K_M | 24.0 | Fits |
| RTX 4080 Super 16GB | 16 GB | Q5_K_M | 20.6 | Too big |
MacBook Pro M4 Pro 48GB | 48 GB | Q5_K_M | 20.2 | Fits |
| RTX 4070 12GB | 12 GB | Q5_K_M | 7.2 | Too big |
| RTX 3060 12GB | 12 GB | Q5_K_M | 4.5 | Too big |
| RTX 4060 8GB | 8 GB | Q5_K_M | 3.2 | Too big |
Estimates for single-stream decoding at Q5_K_M; real tokens/sec varies with prompt length, context, batch size, and runtime build. Prompt processing (prefill) is faster than the decode figures shown here.
How Nous Dolphin 13B (13B params) fits at each quantization level on Intel Arc Pro B60 24GB (24.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 5.1 GB | Low | B66 |
Q3_K_S | 3 | 6.4 GB | Low | B67 |
NVFP4 | 4 | 7.3 GB | Medium | B67 |
Q4_K_M | 4 | 7.9 GB | Medium | B68 |
Q5_K_M | 5 | 9.4 GB | High | B69 |
Q6_K | 6 | 10.7 GB | High | B70 |
Q8_0Best for your GPU | 8 | 13.9 GB | Very High | A71 |
F16 | 16 | 26.7 GB | Maximum | F0 |
Copy-paste commands to run Nous Dolphin 13B on your machine.
Run
docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
--hf-repo "nousresearch/Nous-Dolphin-13B" \
--hf-file "Nous-Dolphin-13B-Q5_K_M.gguf" \
-c 4096 -ngl 99Upgrade options
Adds memory headroom for longer context windows and future model growth.
~$1,599 MSRP
Raises estimated decode speed by about 121%.
Adds memory headroom for longer context windows and future model growth.
~$1,899 MSRP
Yes, Intel Arc Pro B60 24GB can run Nous Dolphin 13B with a B grade (Runs with offload (needs ~0.4 GB host RAM)). Expected decode speed: 18.6 tok/s.
Nous Dolphin 13B (13B parameters) requires approximately 25.2 GB of memory with Q5_K_M quantization.
The recommended quantization for Nous Dolphin 13B is Q5_K_M, which balances quality and memory efficiency.
On Intel Arc Pro B60 24GB, Nous Dolphin 13B achieves approximately 18.6 tokens per second decode speed with a time-to-first-token of 10394ms using Q5_K_M quantization.
For coding workloads, Nous Dolphin 13B on Intel Arc Pro B60 24GB receives a B grade with 18.6 tok/s and 14K context.
On Intel Arc Pro B60 24GB, Nous Dolphin 13B can safely use up to 14K tokens of context. The model's official context limit is 16K, but available memory constrains the safe maximum.
Prefer CUDA if you want the path of least resistance. If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
Often yes, if your goal is the easiest setup and the widest runtime support. Intel can offer attractive memory capacity, but CUDA still tends to win on tooling maturity, guides, kernels, and model coverage for local AI.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/nous-dolphin-13b-on-arc-pro-b60-24gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: