Mixtral-8x7B-Instruct (MoE) VRAM Requirements & GPU Benchmarks
Mixtral-8x7B-Instruct (MoE) (46.7B total, 12.9B active, 32L, 8 KV heads, vocab 32000) requires 27.8 GB in 4-bit Q4_K_M, 46.35 GB in FP8, and 88.47 GB in FP16 at 8K context.
CLOUD-GPU PREISVERGLEICH & MIETEN-VS-KAUFEN RECHNER
Nicht genug lokaler VRAM? Cloud-GPU in 60 Sekunden starten (Mixtral-8x7B-Instruct (MoE): 27.8 GB in 4-bit)
Recommended Cloud Tier: L40S / RTX 6000 Ada (48GB) Β· Full FP16 Needs: 88.47 GB
Amortisationszeit (Kauf vs. Miete): ~32 months (vs $1,699 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 β |
| Vast.ai | $0.29 / hr | $0.95 / hr | $1.79 / hr | $34.80 / mo | Save 50β70% on RTX 4090 / 5090 / A100 rentals | Spot-GPU mieten β |
| Vultr Cloud GPU | β | $1.79 / hr | $2.69 / hr | $214.80 / mo (A100) | $300 Free Credit | $300 Guthaben sichern β |
| DigitalOcean GPU | β | β | $2.99 / hr | $358.80 / mo (H100) | $200 Free Credit (60d) | $200 Startguthaben β |
| GPU Hardware | VRAM | Bandwidth | FP16 (88.47GB) | Q4_K_M (27.8GB) | Decode / Prefill Speed | Report |
|---|---|---|---|---|---|---|
| NVIDIA RTX 5090 | 32 GB | 1792 GB/s | Nein | Ja (87%) | ~158.1 / 4222 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 4090 | 24 GB | 1008 GB/s | Nein | Nein | ~36.1 / 3330 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 5080 | 16 GB | 960 GB/s | Nein | Nein | ~15.9 / 2271 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 4080 Super | 16 GB | 736 GB/s | Nein | Nein | ~15.4 / 2104 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 4070 Ti Super | 16 GB | 672 GB/s | Nein | Nein | ~15.2 / 1778 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 3090 | 24 GB | 936 GB/s | Nein | Nein | ~35.1 / 1431 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 4060 Ti 16GB | 16 GB | 288 GB/s | Nein | Nein | ~12.7 / 889 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 3060 12GB | 12 GB | 360 GB/s | Nein | Nein | ~11.3 / 512 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX 6000 Ada | 48 GB | 960 GB/s | Nein | Ja (58%) | ~84.7 / 3672 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA RTX A6000 | 48 GB | 768 GB/s | Nein | Ja (58%) | ~67.8 / 1560 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA L40S 48GB | 48 GB | 864 GB/s | Nein | Ja (58%) | ~76.2 / 7296 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA A100 80GB SXM | 80 GB | 2039 GB/s | Nein | Ja (35%) | ~179.9 / 6288 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA H100 80GB SXM | 80 GB | 3350 GB/s | Nein | Ja (35%) | ~295.6 / 18000 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA H200 141GB SXM | 141 GB | 4800 GB/s | Ja | Ja (20%) | ~350 / 18000 tok/s | Exakten VRAM & Speed berechnen β |
| NVIDIA Blackwell B200 192GB | 192 GB | 8000 GB/s | Ja | Ja (14%) | ~350 / 18000 tok/s | Exakten VRAM & Speed berechnen β |
| AMD Instinct MI300X 192GB | 192 GB | 5300 GB/s | Ja | Ja (14%) | ~350 / 18000 tok/s | Exakten VRAM & Speed berechnen β |
| 2x NVIDIA RTX 3090 (48GB) | 48 GB | 936 GB/s | Nein | Ja (58%) | ~82.6 / 2862 tok/s | Exakten VRAM & Speed berechnen β |
| 2x NVIDIA RTX 4090 (48GB) | 48 GB | 1008 GB/s | Nein | Ja (58%) | ~88.9 / 6659 tok/s | Exakten VRAM & Speed berechnen β |
| 4x NVIDIA A100 80GB (320GB) | 320 GB | 2039 GB/s | Ja | Ja (9%) | ~179.9 / 18000 tok/s | Exakten VRAM & Speed berechnen β |
| 8x NVIDIA H100 80GB (640GB) | 640 GB | 3350 GB/s | Ja | Ja (4%) | ~295.6 / 18000 tok/s | Exakten VRAM & Speed berechnen β |
| Apple M4 Max (128GB Unified) | 128 GB | 546 GB/s | Ja | Ja (22%) | ~52.7 / 742 tok/s | Exakten VRAM & Speed berechnen β |
| Apple M3 Ultra Mac Studio (512GB) | 512 GB | 819 GB/s | Ja | Ja (5%) | ~79 / 1092 tok/s | Exakten VRAM & Speed berechnen β |
| Apple M4 Pro Mac Mini (64GB) | 64 GB | 273 GB/s | Nein | Ja (43%) | ~26.3 / 371 tok/s | Exakten VRAM & Speed berechnen β |