NVIDIA · Ampere · 2020
NVIDIA RTX 3090
The used-market classic: 24 GB for a few hundred dollars, or pennies per hour rented.
24 GB
GDDR6X
0.936 TB/s
Mem bandwidth
71 TF
FP16 dense
350 W
TDP
$0.18 /hr
From · Vast.ai
Same 24 GB as the 4090 with two-thirds the bandwidth and much less compute. The last consumer card with functional NVLink, which the local-LLM community still exploits for 2-GPU 48 GB rigs.
Where to rent a RTX 3090
On-demand prices per GPU-hour; marketplace rates refresh daily, list prices reviewed 2026-09. Reserved and spot run 30-70% lower. Full breakdown with the used market and rent-vs-buy math: RTX 3090 pricing.
| Provider | $/GPU-hr | Notes |
|---|---|---|
| Vast.ai | $0.18 | Verified-listing rate, refreshed daily |
| RunPod | $0.22 | Community Cloud |
Specifications in context
Bars scaled against the best value in the whole catalog (B200 / MI325X era).
| Architecture | Ampere (2020) |
|---|---|
| Memory | 24 GB GDDR6X, 0.936 TB/s |
| Interconnect | PCIe Gen4 (NVLink bridge) |
| Form factor | Consumer card |
| Partitioning | No MIG (time-slicing / vGPU only) |
What fits on one RTX 3090
Weights + ~2 GB runtime overhead against 22 GB usable VRAM. Longer contexts and bigger batches need more, check the VRAM calculator.
| Model | Params | Highest precision that fits |
|---|---|---|
| Llama 3.2 1B | 1.24B | FP16 |
| Llama 3.2 3B | 3.21B | FP16 |
| Llama 3.1 8B | 8.03B | FP16 |
| Qwen2.5 7B | 7.62B | FP16 |
| Qwen2.5 14B | 14.8B | 8-bit |
| Mistral 7B | 7.25B | FP16 |
| Gemma 2 9B | 9.24B | 8-bit |
| Gemma 2 27B | 27.2B | 4-bit |
| Phi-4 14B | 14.7B | 8-bit |
| gpt-oss-20b | 20.9B | 4-bit |
Best for
- Entry local LLM rigs
- Cheapest 24 GB rentals anywhere
- 2×3090 NVLink 48 GB builds