✅ EJECUTA FP16 NATIVO Y TODAS LAS CUANTIZACIONES
Llama-3.2-3B-Instruct on NVIDIA L40S 48GB (48GB VRAM)
In 4-bit GGUF Q4_K_M, Llama-3.2-3B-Instruct consumes 3.33 GB (1.96 GB weights + 0.88 GB KV cache at 8K context), achieving ~306.4 tokens/sec decode and ~18000 tokens/sec prefill.
🎛️ Ajustador Interactivo de VRAM, Contexto y Offload
📊 Telemetría de Memoria y Rendimiento en Tiempo Real
3.33 GB / 48 GB
🟦 Pesos del Modelo: 1.96 GB
🟪 Caché KV: 0.88 GB
🟨 Buffer del Motor: 0.49 GB
VRAM de GPU Requerida3.33 GB
RAM de Sistema Requerida (CPU)0.00 GB
VRAM Disponible en Hardware48 GB (1x NVIDIA L40S 48GB)
Margen / Déficit de VRAM44.67 GB
Velocidad de Generación (Decode)~306.4 tokens/sec
Velocidad de Procesamiento (Prefill)~18000 tokens/sec
Tiempo al Primer Token (Prompt 4K)~228 ms
Topología TransformerDense-GQA (28L, 8/24 KV/Q heads)
🚀 Comando de Lanzamiento Listo para Copiar
ollama run llama3.2:3b # Or llama.cpp server (8K context, Flash Attention): llama-server -hf meta-llama/Llama-3.2-3B-Instruct-GGUF -c 8192 -ngl 99 -fa
COMPARADOR CLOUD GPU Y CALCULADORA ALQUILAR VS COMPRAR
¿No tienes suficiente VRAM? Alquila una GPU en la nube en 60 segundos (Llama-3.2-3B-Instruct: 3.33 GB in 4-bit)
Recommended Cloud Tier: RTX 4090 (24GB) · Full FP16 Needs: 7.78 GB
Meses para amortizar compra vs alquiler: ~32 months (vs $8,500 hardware)
| Cloud Provider | RTX 4090 (24GB) | A100 (80GB) | H100 (80GB) | Est. Monthly Cost (4h/day) | Free Credit / Trial | Deploy |
|---|---|---|---|---|---|---|
| RunPod | $0.44 / hr | $1.39 / hr | $2.39 / hr | $52.80 / mo | Instant PyTorch / vLLM Templates + Serverless GPU | Desplegar en RunPod → |
| Vast.ai | $0.29 / hr | $0.95 / hr | $1.79 / hr | $34.80 / mo | Save 50–70% on RTX 4090 / 5090 / A100 rentals | Alquilar GPU Spot → |
| Vultr Cloud GPU | — | $1.79 / hr | $2.69 / hr | $214.80 / mo (A100) | $300 Free Credit | Reclamar $300 Gratis → |
| DigitalOcean GPU | — | — | $2.99 / hr | $358.80 / mo (H100) | $200 Free Credit (60d) | Obtener $200 Gratis → |
Desglose Completo por Formato de Cuantización: Llama-3.2-3B-Instruct × NVIDIA L40S 48GB
| Quantization Format | Core / Emb BPW | Weights (GiB) | KV Cache (8K) | Total VRAM | 1x NVIDIA L40S 48GB (48GB) | Decode / Prefill | Min GPUs |
|---|---|---|---|---|---|---|---|
| FP16 / BF16 (Unquantized) | 16 / 16 | 5.98 GB | 0.88 GB | 7.34 GB | Sí (15%) | ~92.5 / 18000 tok/s | 1x |
| FP8 (W8A8 / E4M3) | 8.25 / 8.5 | 3.11 GB | 0.88 GB | 4.47 GB | Sí (9%) | ~179.4 / 18000 tok/s | 1x |
| GGUF Q8_0 (8.50 bpw) | 8.5 / 8.5 | 3.18 GB | 0.88 GB | 4.54 GB | Sí (9%) | ~174.1 / 18000 tok/s | 1x |
| GGUF Q6_K (6.56 bpw) | 6.56 / 6.56 | 2.45 GB | 0.88 GB | 3.81 GB | Sí (8%) | ~225.6 / 18000 tok/s | 1x |
| GGUF Q5_K_M (5.69 bpw) | 5.69 / 6.56 | 2.21 GB | 0.88 GB | 3.57 GB | Sí (7%) | ~260.1 / 18000 tok/s | 1x |
| GGUF Q4_K_M / AWQ (4.83 bpw) | 4.83 / 6.56 | 1.96 GB | 0.88 GB | 3.33 GB | Sí (7%) | ~306.4 / 18000 tok/s | 1x |
| GGUF IQ4_XS (4.25 bpw) | 4.25 / 5.5 | 1.7 GB | 0.88 GB | 3.06 GB | Sí (6%) | ~348.2 / 18000 tok/s | 1x |
| GGUF Q3_K_M (3.91 bpw) | 3.91 / 5 | 1.56 GB | 0.88 GB | 2.92 GB | Sí (6%) | ~350 / 18000 tok/s | 1x |
Probar este Modelo en Otras GPUs
Llama-3.2-3B-Instruct on NVIDIA RTX 5090 (32GB)Llama-3.2-3B-Instruct on NVIDIA RTX 4090 (24GB)Llama-3.2-3B-Instruct on NVIDIA RTX 5080 (16GB)Llama-3.2-3B-Instruct on NVIDIA RTX 4080 Super (16GB)Llama-3.2-3B-Instruct on NVIDIA RTX 4070 Ti Super (16GB)Llama-3.2-3B-Instruct on NVIDIA RTX 3090 (24GB)Llama-3.2-3B-Instruct on NVIDIA RTX 4060 Ti 16GB (16GB)Llama-3.2-3B-Instruct on NVIDIA RTX 3060 12GB (12GB)Llama-3.2-3B-Instruct on NVIDIA RTX 6000 Ada (48GB)Llama-3.2-3B-Instruct on NVIDIA RTX A6000 (48GB)
Otros Modelos en esta GPU
DeepSeek-R1 671B (MoE) on NVIDIA L40S 48GBDeepSeek-V3-0324 671B (MoE) on NVIDIA L40S 48GBDeepSeek-R1-Distill-Llama-70B on NVIDIA L40S 48GBDeepSeek-R1-Distill-Qwen-32B on NVIDIA L40S 48GBDeepSeek-R1-Distill-Qwen-14B on NVIDIA L40S 48GBDeepSeek-R1-Distill-Llama-8B on NVIDIA L40S 48GBDeepSeek-R1-Distill-Qwen-7B on NVIDIA L40S 48GBQwen3-235B-A22B (MoE) on NVIDIA L40S 48GBQwen3-30B-A3B (MoE) on NVIDIA L40S 48GBQwen3-32B on NVIDIA L40S 48GB