Can SenseNova Loop 9B Preview run on the RTX 3080 10GB?
Yes. At 8K context the best fit is Q6_K, using about 8.0 GB of the 9.5 GB usable on the RTX 3080 10GB. Estimated speed ceiling: ~103 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | 128K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|---|
| 2-bit (Q2_K) | 4.4 GB Fits | 4.5 GB Fits | 5.2 GB Fits | 8.2 GB Fits | 210 |
| 3-bit (Q3_K_M) | 5.4 GB Fits | 5.5 GB Fits | 6.3 GB Fits | 9.3 GB Tight | 161 |
| 4-bit (Q4_K_M) • | 6.2 GB Fits | 6.4 GB Fits | 7.1 GB Fits | 10.1 GB No | 135 |
| 5-bit (Q5_K_M) • | 7.0 GB Fits | 7.2 GB Fits | 7.9 GB Fits | 10.9 GB No | 117 |
| 6-bit (Q6_K) • | 7.9 GB Fits | 8.0 GB Fits | 8.7 GB Tight | 11.7 GB No | 103 |
| 8-bit (Q8_0) • | 9.9 GB No | 10.0 GB No | 10.8 GB No | 13.8 GB No | — |
| 16-bit (F16/BF16) • | 17.7 GB No | 17.8 GB No | 18.6 GB No | 21.6 GB No | — |
• real GGUF file size from AtomicChat/SenseNova-Loop-9B-Preview-GGUF. Speed is a bandwidth ceiling (760 GB/s), not a benchmark. Methodology.
More
SenseNova Loop 9B Preview on every GPU and Mac · Every model that runs on the RTX 3080 10GB · Other models that fit 12 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (sensenova/SenseNova-Loop-9B-Preview), NVIDIA specs. Data as of 2026-11-17.