← Back to community benchmarks
NVIDIA-Nemotron-3-Nano-30B-A3B
Performance
128k
tokens
3,168
PP tok/s
60.5
TG tok/s
41370
TTFT (ms)
37.4
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
131,072
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,382 | 67.8 | 34.6 GB | view |
| 4k | 1,863 | 66.4 | 35.7 GB | view |
| 8k | 1,908 | 64.4 | 35.8 GB | view |
| 16k | 4,291 | 66.7 | 35.6 GB | view |
| 32k | 4,428 | 66.1 | 35.7 GB | view |
| 64k | 4,007 | 62.5 | 35.7 GB | view |
| 128k | 3,168 | 60.5 | 37.4 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 67.8 | 1.00× |
| 2× | 92.9 | 1.37× |
| 4× | 114.9 | 1.69× |
| 8× | 128.2 | 1.89× |