← Back to community benchmarks
NVIDIA-Nemotron-3-Nano-30B-A3B
Performance
8k
tokens
3,128
PP tok/s
121.1
TG tok/s
2619
TTFT (ms)
27.2
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.6
macOS
macOS 26.3.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,478 | 131.6 | 26.2 GB | view |
| 4k | 2,840 | 123.1 | 27.2 GB | view |
| 8k | 3,128 | 121.1 | 27.2 GB | current |
| 16k | 3,286 | 116.3 | 27.3 GB | view |
| 32k | 3,413 | 98.9 | 27.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 131.6 | 1.00× |
| 2× | 183.5 | 1.39× |
| 4× | 253.2 | 1.92× |
| 8× | 245.7 | 1.87× |