✅ LÄUFT NATIV IN FP16 & ALLEN QUANTISIERUNGEN
Qwen3-8B on NVIDIA RTX 3090 (24GB VRAM)
In 4-bit GGUF Q4_K_M, Qwen3-8B consumes 6.47 GB (4.86 GB weights + 1.13 GB KV cache at 8K context), achieving ~130.1 tokens/sec decode and ~2254 tokens/sec prefill.
🎛️ Interaktiver VRAM-, Kontext- & Offload-Konfigurator
📊 Echtzeit-Speicher- & Performance-Telemetrie
6.47 GB / 24 GB
🟦 Modellgewichte: 4.86 GB
🟪 KV-Cache: 1.13 GB
🟨 Engine-Puffer: 0.49 GB
Benötigter GPU-VRAM6.47 GB
Benötigter System-RAM (CPU)0.00 GB
Verfügbarer Hardware-VRAM24 GB (1x NVIDIA RTX 3090)
VRAM-Reserve / Defizit17.53 GB
Geschätzte Decode-Geschwindigkeit~130.1 tokens/sec
Geschätzte Prefill-Geschwindigkeit~2254 tokens/sec
Time-to-First-Token (4K Prompt)~1817 ms
Transformer-TopologieDense-GQA (36L, 8/32 KV/Q heads)
🚀 Startbefehl (vLLM / llama.cpp / Ollama)
ollama run qwen3:8b # Or llama.cpp server (8K context, Flash Attention): llama-server -hf Qwen/Qwen3-8B-GGUF -c 8192 -ngl 99 -fa
CLOUD-GPU PREISVERGLEICH & MIETEN-VS-KAUFEN RECHNER
Nicht genug lokaler VRAM? Cloud-GPU in 60 Sekunden starten (Qwen3-8B: 6.47 GB in 4-bit)
Recommended Cloud Tier: RTX 4090 (24GB) · Full FP16 Needs: 17.31 GB
Amortisationszeit (Kauf vs. Miete): ~32 months (vs $750 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 | Auf RunPod starten → |
| Vast.ai | $0.29 / hr | $0.95 / hr | $1.79 / hr | $34.80 / mo | Save 50–70% on RTX 4090 / 5090 / A100 rentals | Spot-GPU mieten → |
| Vultr Cloud GPU | — | $1.79 / hr | $2.69 / hr | $214.80 / mo (A100) | $300 Free Credit | $300 Guthaben sichern → |
| DigitalOcean GPU | — | — | $2.99 / hr | $358.80 / mo (H100) | $200 Free Credit (60d) | $200 Startguthaben → |
Alle Quantisierungsstufen im Detail: Qwen3-8B × NVIDIA RTX 3090
| Quantization Format | Core / Emb BPW | Weights (GiB) | KV Cache (8K) | Total VRAM | 1x NVIDIA RTX 3090 (24GB) | Decode / Prefill | Min GPUs |
|---|---|---|---|---|---|---|---|
| FP16 / BF16 (Unquantized) | 16 / 16 | 15.26 GB | 1.13 GB | 16.87 GB | Ja (70%) | ~39.3 / 2254 tok/s | 1x |
| FP8 (W8A8 / E4M3) | 8.25 / 8.5 | 7.9 GB | 1.13 GB | 9.52 GB | Ja (40%) | ~76.2 / 2254 tok/s | 1x |
| GGUF Q8_0 (8.50 bpw) | 8.5 / 8.5 | 8.1 GB | 1.13 GB | 9.72 GB | Ja (41%) | ~73.9 / 2254 tok/s | 1x |
| GGUF Q6_K (6.56 bpw) | 6.56 / 6.56 | 6.25 GB | 1.13 GB | 7.87 GB | Ja (33%) | ~95.8 / 2254 tok/s | 1x |
| GGUF Q5_K_M (5.69 bpw) | 5.69 / 6.56 | 5.55 GB | 1.13 GB | 7.16 GB | Ja (30%) | ~110.4 / 2254 tok/s | 1x |
| GGUF Q4_K_M / AWQ (4.83 bpw) | 4.83 / 6.56 | 4.86 GB | 1.13 GB | 6.47 GB | Ja (27%) | ~130.1 / 2254 tok/s | 1x |
| GGUF IQ4_XS (4.25 bpw) | 4.25 / 5.5 | 4.23 GB | 1.13 GB | 5.85 GB | Ja (24%) | ~147.8 / 2254 tok/s | 1x |
| GGUF Q3_K_M (3.91 bpw) | 3.91 / 5 | 3.89 GB | 1.13 GB | 5.5 GB | Ja (23%) | ~160.7 / 2254 tok/s | 1x |
Dieses Modell auf anderen GPUs testen
Qwen3-8B on NVIDIA RTX 5090 (32GB)Qwen3-8B on NVIDIA RTX 4090 (24GB)Qwen3-8B on NVIDIA RTX 5080 (16GB)Qwen3-8B on NVIDIA RTX 4080 Super (16GB)Qwen3-8B on NVIDIA RTX 4070 Ti Super (16GB)Qwen3-8B on NVIDIA RTX 4060 Ti 16GB (16GB)Qwen3-8B on NVIDIA RTX 3060 12GB (12GB)Qwen3-8B on NVIDIA RTX 6000 Ada (48GB)Qwen3-8B on NVIDIA RTX A6000 (48GB)Qwen3-8B on NVIDIA L40S 48GB (48GB)
Weitere Modelle auf dieser GPU
DeepSeek-R1 671B (MoE) on NVIDIA RTX 3090DeepSeek-V3-0324 671B (MoE) on NVIDIA RTX 3090DeepSeek-R1-Distill-Llama-70B on NVIDIA RTX 3090DeepSeek-R1-Distill-Qwen-32B on NVIDIA RTX 3090DeepSeek-R1-Distill-Qwen-14B on NVIDIA RTX 3090DeepSeek-R1-Distill-Llama-8B on NVIDIA RTX 3090DeepSeek-R1-Distill-Qwen-7B on NVIDIA RTX 3090Qwen3-235B-A22B (MoE) on NVIDIA RTX 3090Qwen3-30B-A3B (MoE) on NVIDIA RTX 3090Qwen3-32B on NVIDIA RTX 3090