← CanIRunLocal · All hardware · All models
NVIDIA · 120B parameters (12B active) · chat, reasoning, coding, vision.
Ollama weight files: Q4_K_M 90 GB (nemotron-3.5-super:120b-a12b-q4_K_M); Q8_0 133 GB (nemotron-3.5-super:120b-a12b-q8_0).
Yes on 16 machines in this list. The smallest is M3 Max 128GB: Q4_K_M, about 92.3 GB, roughly 18 tokens per second, via Ollama. ollama run nemotron-3.5-super:120b-a12b-q4_K_M
ollama run nemotron-3.5-super:120b-a12b-q4_K_Mollama run nemotron-3.5-super:120b-a12b-q4_K_Mollama run nemotron-3.5-super:120b-a12b-q8_0ollama run nemotron-3.5-super:120b-a12b-q4_K_M| Hardware | Runtime | Quant | Memory | Fit | Speed | Command |
|---|---|---|---|---|---|---|
| M3 Max 128GB | Ollama | Q4_K_M | 92.3 GB | fits in fast memory | ~18 tok/s | ollama run nemotron-3.5-super:120b-a12b-q4_K_M |
| M4 Max 128GB | Ollama | Q4_K_M | 92.3 GB | fits in fast memory | ~24 tok/s | ollama run nemotron-3.5-super:120b-a12b-q4_K_M |
| Mac Studio M4 Max (40c GPU) 128GB | Ollama | Q4_K_M | 92.3 GB | fits in fast memory | ~24 tok/s | ollama run nemotron-3.5-super:120b-a12b-q4_K_M |
| Mac Studio M5 Max (40c GPU) 128GB | Ollama | Q4_K_M | 92.3 GB | fits in fast memory | ~27 tok/s | ollama run nemotron-3.5-super:120b-a12b-q4_K_M |
| DGX Spark 128GB (NVIDIA, ASUS, Dell, Lenovo, HP, MSI, Acer, Gigabyte) | Ollama | Q4_K_M | 92.3 GB | fits in fast memory | ~17 tok/s | ollama run nemotron-3.5-super:120b-a12b-q4_K_M |
| M2 Ultra 192GB | Ollama | Q8_0 | 135.3 GB | fits in fast memory | ~24 tok/s | ollama run nemotron-3.5-super:120b-a12b-q8_0 |
| 2x Mac Studio M5 Max (40c GPU) 128GB | llama.cpp | Q8_0 GGUF | 134 GB | fits in fast memory | ~16 tok/s | llama-server -hf unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| Mac Studio M3 Ultra 256GB | Ollama | Q8_0 | 135.3 GB | fits in fast memory | ~25 tok/s | ollama run nemotron-3.5-super:120b-a12b-q8_0 |
| Mac Studio M5 Ultra (64c GPU) 256GB | Ollama | Q8_0 | 135.3 GB | fits in fast memory | ~36 tok/s | ollama run nemotron-3.5-super:120b-a12b-q8_0 |
| Mac Studio M5 Ultra (80c GPU) 256GB | Ollama | Q8_0 | 135.3 GB | fits in fast memory | ~36 tok/s | ollama run nemotron-3.5-super:120b-a12b-q8_0 |
| 2x DGX Spark 128GB | llama.cpp | Q8_0 GGUF | 134 GB | fits in fast memory | ~10 tok/s | llama-server -hf unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| 4x Mac Studio M5 Max (40c GPU) 128GB | llama.cpp | Q8_0 GGUF | 134 GB | fits in fast memory | ~14 tok/s | llama-server -hf unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF:Q8_0 --rpc node2:50052,node3:50052,node4:50052 -c 8192 # run rpc-server on each other node |
| 2x Mac Studio M5 Ultra (80c GPU) 256GB | llama.cpp | Q8_0 GGUF | 134 GB | fits in fast memory | ~31 tok/s | llama-server -hf unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| Mac Studio M3 Ultra (80c GPU) 512GB | Ollama | Q8_0 | 135.3 GB | fits in fast memory | ~25 tok/s | ollama run nemotron-3.5-super:120b-a12b-q8_0 |
| 4x DGX Spark 128GB | llama.cpp | Q8_0 GGUF | 134 GB | fits in fast memory | ~9 tok/s | llama-server -hf unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF:Q8_0 --rpc node2:50052,node3:50052,node4:50052 -c 8192 # run rpc-server on each other node |
| 4x Mac Studio M5 Ultra (80c GPU) 256GB | llama.cpp | Q8_0 GGUF | 134 GB | fits in fast memory | ~27 tok/s | llama-server -hf unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF:Q8_0 --rpc node2:50052,node3:50052,node4:50052 -c 8192 # run rpc-server on each other node |
Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-16.
A project by OnDevice Labs · llms.txt