Can the RTX 3050 OEM run Qwen3.8-27B-OBLITERATED?
No. Qwen3.8-27B-OBLITERATED needs about 17.8 GB at Q4_K_M (weights plus 1.8 GB runtime overhead), and the RTX 3050 OEM has 8 GB, so it won't fit without offloading layers to system RAM, which is far slower.
Every quantization
Qwen3.8-27B-OBLITERATED (27.8B) on 8 GB
| Quant | Weights | Needs | Fit | Est. speed |
|---|---|---|---|---|
| Q4_K_M | 16 GB | 17.8 GB | Too big | - |
| Q5_K_M | 20 GB | 21.8 GB | Too big | - |
| Q8_0 | 30 GB | 31.8 GB | Too big | - |
| FP16 | 56 GB | 57.8 GB | Too big | - |
Estimates. Needs = weights + 1.8 GB for KV cache and runtime at modest context. Speed is single-stream decode scaled from the RTX 3050 OEM's 224 GB/s bandwidth (Estimated from memory bandwidth (llama.cpp decode)). Long contexts need more memory and run slower.
What to do instead
Two ways forward
Cheapest cards that run it
- NVIDIA Tesla P40 - 24 GB, ~$159
- AMD Radeon RX 7900 XT - 20 GB, ~$699
- NVIDIA RTX 3090 - 24 GB, ~$750
- AMD Radeon RX 7900 XTX - 24 GB, ~$849
Models that fit the RTX 3050 OEM
- Qwen3.5 4B - 4.7B, 3 GB at Q4
- Qwen3-4B-Instruct-2507 - 4B, 2 GB at Q4
- Phi-4 Mini - 3B, 2.5 GB at Q4
Compare
Same model, other hardware - and other models on this card
Qwen3.8-27B-OBLITERATED on other hardware
- RTX 5090~64 tok/s
- RTX 5090 D~64 tok/s
- RTX 6000D~56 tok/s
- RTX 4090~39 tok/s
- RTX 3090~34 tok/s
- Radeon RX 7900 XTX~34 tok/s
- MacBook Pro 16" M4 Max~20 tok/s
- Arc Pro B60~16 tok/s
Also runs on the RTX 3050 OEM
- Qwen3.5 4B~43 tok/s
- Qwen3-4B-Instruct-2507~64 tok/s
- Phi-4 Mini~51 tok/s
Frequently asked
Quick answers
Can the GeForce RTX 3050 OEM run Qwen3.8-27B-OBLITERATED?
No. Qwen3.8-27B-OBLITERATED needs about 17.8 GB at Q4_K_M (weights plus 1.8 GB runtime overhead), and the RTX 3050 OEM has 8 GB, so it won't fit without offloading layers to system RAM, which is far slower.
How much VRAM does Qwen3.8-27B-OBLITERATED need?
About 16 GB of weights at Q4_K_M, 30 GB at Q8_0 and 56 GB at FP16. Add roughly 1.8 GB for the KV cache and runtime at modest context lengths.
How fast is Qwen3.8-27B-OBLITERATED on the RTX 3050 OEM?
It doesn't fit in the RTX 3050 OEM's memory, so speed depends on how many layers you offload to system RAM - typically single-digit tokens per second or lower.