← Back to community benchmarks
NVIDIA-Nemotron-3-Nano-30B-A3B
Performance
4k
tokens
831.0
PP tok/s
58.0
TG tok/s
4929
TTFT (ms)
34.6
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.8
macOS
macOS 15.7.4
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 673.7 | 59.7 | 33.5 GB | view |
| 4k | 831.0 | 58.0 | 34.6 GB | current |
| 8k | 842.1 | 54.8 | 34.6 GB | view |
| 16k | 808.3 | 52.3 | 34.7 GB | view |
| 32k | 772.9 | 50.1 | 34.7 GB | view |
| 64k | 688.4 | 43.2 | 34.9 GB | view |
| 128k | 496.7 | 31.9 | 35.3 GB | view |
| 195k | 379.5 | 26.8 | 35.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 59.7 | 1.00× |
| 2× | 74.7 | 1.25× |
| 4× | 83.0 | 1.39× |
| 8× | 92.0 | 1.54× |