▸ DEVICE UNDER TEST
NVIDIA Quadro FX 3800M — 1 GB VRAM.
▸ QUADRO FX 3800M SPEC
- BRAND
- NVIDIA
- VRAM
- 1 GB GDDR3
- BANDWIDTH
- 64 GB/s
- FP16 COMPUTE
- 0.4 TFLOPS
- FP32 COMPUTE
- 0.4 TFLOPS
- CUDA CORES
- 128
- TDP
- 100 W
- ARCHITECTURE
- Tesla
▸ AI CAPABILITY
23/ 428 models @ Q4
With 1 GB VRAM and 64 GB/s bandwidth, this GPU handles models up to 0.6B parameters.
Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~7 tok/s.
§ 01TOP MODELS FOR QUADRO FX 3800M
23 FIT · SHOWING 20| MODEL | SIZE | VRAM Q4 | TOK/S | AVG |
|---|---|---|---|---|
| Qwen3 0.6B | 0.6B | 0.9 GB | 85 | 19.1 |
| Qwen3-Embedding 0.6B | 0.6B | 0.9 GB | 85 | — |
| Falcon-H1R Tiny 0.6B | 0.6B | 0.9 GB | 85 | — |
| Falcon Perception 0.6B | 0.6B | 0.9 GB | 85 | — |
| BGE-M3 | 0.568B | 0.8 GB | 90 | 63.0 |
| Snowflake Arctic Embed L v2.0 | 0.568B | 0.8 GB | 90 | — |
| Falcon-H1 0.5B | 0.5B | 0.8 GB | 102 | 41.7 |
| Qwen 1.5 0.5B | 0.5B | 0.8 GB | 102 | 9.9 |
| Qwen 2.5 0.5B | 0.5B | 0.8 GB | 102 | 19.4 |
| SmolVLM 500M | 0.5B | 0.8 GB | 102 | — |
| SmolLM2 360M | 0.36B | 0.7 GB | 142 | 8.2 |
| LFM2 350M | 0.35B | 0.7 GB | 146 | 46.3 |
| GPT-2 Medium 345M | 0.345B | 0.7 GB | 148 | 5.9 |
| bge-large-en-v1.5 335M | 0.335B | 0.7 GB | 153 | 62.3 |
| mxbai-embed-large-v1 | 0.335B | 0.7 GB | 153 | 64.7 |
| Snowflake Arctic Embed L | 0.335B | 0.7 GB | 153 | 56.0 |
| Snowflake Arctic Embed M v2.0 | 0.305B | 0.7 GB | 168 | — |
| Gemma 3 270M | 0.27B | 0.7 GB | 190 | 12.6 |
| SmolVLM 256M | 0.256B | 0.6 GB | 200 | 28.3 |
| SmolLM2 135M | 0.135B | 0.6 GB | 379 | 7.0 |