▸ DEVICE UNDER TEST
NVIDIA RTX 3500 Embedded Ada Generation — 12 GB VRAM.
▸ RTX 3500 EMBEDDED ADA GENERATION SPEC
- BRAND
- NVIDIA
- VRAM
- 12 GB GDDR6
- BANDWIDTH
- 432 GB/s
- FP16 COMPUTE
- 23 TFLOPS
- FP32 COMPUTE
- 23 TFLOPS
- CUDA CORES
- 5,120
- TENSOR CORES
- 160
- TDP
- 100 W
- ARCHITECTURE
- Ada Lovelace
▸ AI CAPABILITY
250/ 428 models @ Q4
With 12 GB VRAM and 432 GB/s bandwidth, this GPU handles models up to 16.8B parameters.
Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~49 tok/s.
§ 01TOP MODELS FOR RTX 3500 EMBEDDED ADA GENERATION
250 FIT · SHOWING 20| MODEL | SIZE | VRAM Q4 | TOK/S | AVG |
|---|---|---|---|---|
| Ling-lite 16.8B | 16.8B | 10.8 GB | 144 | — |
| DeepSeek V2 Lite 16B | 16B | 10.3 GB | 144 | 38.0 |
| DeepSeek-Coder-V2-Lite 15.7B | 15.7B | 10.1 GB | 144 | 43.0 |
| DeepSeek-VL2 Small 16B | 15.7B | 10.1 GB | 144 | 43.1 |
| StarCoder 15B | 15.5B | 10.0 GB | 22 | 21.0 |
| StarCoder2 15B | 15B | 9.7 GB | 23 | 26.5 |
| Phi-4-reasoning-vision 15B | 15B | 9.7 GB | 23 | 42.8 |
| DeepSeek R1 Distill Qwen 14B | 14.8B | 9.5 GB | 23 | 43.9 |
| DeepCoder 14B | 14.8B | 9.5 GB | 23 | 38.7 |
| Qwen2.5-Coder-14B | 14.8B | 9.5 GB | 23 | 41.3 |
| Qwen2.5-14B | 14.8B | 9.5 GB | 23 | 41.3 |
| Qwen3 14B | 14.8B | 9.5 GB | 23 | 45.7 |
| Ministral 3 14B | 14B | 9.0 GB | 25 | 25.9 |
| Phi-3-medium-14b | 14B | 9.0 GB | 25 | 33.7 |
| phi-4 14B | 14B | 9.0 GB | 25 | 33.7 |
| Phi-4-reasoning 14B | 14B | 9.0 GB | 25 | 33.7 |
| Phi-4-multimodal 14B | 14B | 9.0 GB | 25 | 42.0 |
| Qwen 1.5 14B | 14B | 9.0 GB | 25 | 41.3 |
| InternVL3 14B | 14B | 9.0 GB | 25 | 38.1 |
| LLaVA-1.5 13B | 13.1B | 8.5 GB | 26 | 52.1 |