← Back to community benchmarks
NVIDIA-Nemotron-3-Nano-30B-A3B
Performance
8k
tokens
1,906
PP tok/s
61.8
TG tok/s
4298
TTFT (ms)
45.8
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.3.8rc1
macOS
macOS 26.4.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,339 | 65.1 | 44.7 GB | view |
| 4k | 1,848 | 65.9 | 45.8 GB | view |
| 8k | 1,906 | 61.8 | 45.8 GB | current |
| 16k | 1,978 | 64.9 | 45.7 GB | view |
| 32k | 1,829 | 65.1 | 46.6 GB | view |
| 64k | 1,444 | 58.8 | 50.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 65.1 | 1.00× |
| 2× | 87.4 | 1.34× |
| 4× | 112.1 | 1.72× |
| 8× | 126.4 | 1.94× |