Home / Multi-GPU
llama.cpp · homogeneous 2–4 GPUs · not a PC builder
Can this model run across multiple GPUs?
Phase 8 answers memory feasibility for identical cards under llama.cpp. Default split is layer (pipeline parallel: each GPU owns a slice of layers; KV stays with those layers). This is not 2 × VRAM = pool, and it is not a complete PC build.
Cheapest current GPU-only multi-GPU configuration in RigForAI
Homogeneous llama.cpp layer configs that fully fit, among cards with a fresh matched Amazon offer. Cost = GPU count × current per-card price. This is not a complete build and does not prove that N units are in stock.
| # | Configuration | Per-GPU peak | GPU-only cost | Winning SKU |
|---|---|---|---|---|
| 1 | 4× NVIDIA Quadro P6000 24 GB | 19.4 GB | $2587.84 4 × $646.96 Amazon | HP NVIDIA Quadro P6000 (24GB) Graphics Card ASIN B01NALQ0PB |
| 2 | 4× AMD Radeon RX 7900 XTX 24 GB | 19.4 GB | $4239.96 4 × $1059.99 Amazon | GIGABYTE GAMING Radeon RX 7900 XTX OC 24G AMD 24 GB GDDR6 ASIN B0BRBXF3MW |
| 3 | 4× NVIDIA RTX 3090 24 GB | 19.4 GB | $5388.88 4 × $1347.22 Amazon | GIGABYTE EAGLE GV-N3090EAGLE OC-24GD graphics card NVIDIA GeForce RTX 3090 24 GB GDDR6X ASIN B08HJPDJTY |
| 4 | 4× AMD Radeon RX 7900 XT 24 GB | 19.4 GB | $6396.00 4 × $1599.00 Amazon | MSI GAMING Radeon RX 7900 XTX Trio Classic 24G AMD Radeon RX 7900 XT 24 GB GDDR6 ASIN B0BSN878GR |
| 5 | 4× NVIDIA RTX 3090 Ti 24 GB | 19.4 GB | $8840.00 4 × $2210.00 Amazon | MSI SUPRIM GeForce RTX 3090 Ti X 24G NVIDIA 24 GB GDDR6X ASIN B09Q31NH4M |
| 6 | 4× NVIDIA RTX 4090 24 GB | 19.4 GB | $10319.96 4 × $2579.99 Amazon | GIGABYTE AERO GeForce RTX 4090 OC 24G NVIDIA 24 GB GDDR6X ASIN B0BXMP9QVF |
| 7 | 4× NVIDIA RTX A5000 24 GB | 19.4 GB | $11040.00 4 × $2760.00 Amazon | Lenovo 4X61D97085 graphics card NVIDIA RTX A5000 24 GB GDDR6 ASIN B09PJSBFMG |
| 8 | 3× NVIDIA RTX 5000 32 GB | 25.7 GB | $13439.88 3 × $4479.96 Amazon | Lenovo NVIDIA RTX 5000 Ada 32 GB GDDR6 ASIN B0FGQG418C |
| 9 | 4× NVIDIA Quadro M6000 24 GB | 19.4 GB | $13992.08 4 × $3498.02 Amazon | HP NVIDIA Quadro M6000 24GB Graphics Card ASIN B01EL0OQKE |
| 10 | 3× NVIDIA RTX 5090 32 GB | 25.7 GB | $14399.97 3 × $4799.99 Amazon | GIGABYTE GeForce RTX 5090 WINDFORCE OC 32G Graphics Card ASIN B0DT7GMXHB |
| 11 | 4× NVIDIA RTX 5000 32 GB | 19.4 GB | $17919.84 4 × $4479.96 Amazon | Lenovo NVIDIA RTX 5000 Ada 32 GB GDDR6 ASIN B0FGQG418C |
| 12 | 4× NVIDIA RTX 5090 32 GB | 19.4 GB | $19199.96 4 × $4799.99 Amazon | GIGABYTE GeForce RTX 5090 WINDFORCE OC 32G Graphics Card ASIN B0DT7GMXHB |
| 13 | 2× NVIDIA RTX 6000 48 GB | 38.1 GB | $41776.00 2 × $20888.00 Amazon | HP NVIDIA RTX 6000 Ada 48 GB 4DP Graphics ASIN B0CTP2HHBN |
| 14 | 3× NVIDIA RTX 6000 48 GB | 25.7 GB | $62664.00 3 × $20888.00 Amazon | HP NVIDIA RTX 6000 Ada 48 GB 4DP Graphics ASIN B0CTP2HHBN |
| 15 | 4× NVIDIA RTX 6000 48 GB | 19.4 GB | $83552.00 4 × $20888.00 Amazon | HP NVIDIA RTX 6000 Ada 48 GB 4DP Graphics ASIN B0CTP2HHBN |
Feasible — current price unavailable
- 4× NVIDIA RTX 4500 24 GB
- 4× NVIDIA Tesla K80 24 GB
- 4× NVIDIA Tesla M40 24 GB
- 4× NVIDIA RTX 6000 24 GB
- 4× NVIDIA RTX A5500 24 GB
- 3× NVIDIA Tesla M10 32 GB
- 4× NVIDIA Tesla M10 32 GB
- 3× NVIDIA Quadro GV100 32 GB
- 4× NVIDIA Quadro GV100 32 GB
- 2× NVIDIA RTX 8000 48 GB
- 3× NVIDIA RTX 8000 48 GB
- 4× NVIDIA RTX 8000 48 GB
Query combinations are not indexed. Source: llama.cpp multi-GPU documentation.