HARDWARE BENCHMARK & VRAM ENVELOPE โข MULTI-GPU RIG48 GB VRAM โข 936 GB/s
What Models Can Run on 2x NVIDIA RTX 3090 Dual Rig (48GB Total VRAM)?
48 GB VRAM936 GB/s Bandwidth700W TDPMSRP: $1,550Cloud: $0.78/hr on Vast.ai
2x NVIDIA RTX 3090 Dual Rig (48GB Total VRAM) provides 48 GB of usable memory with 936 GB/s memory bandwidth at 700W TDP. Best price-to-VRAM homelab rig for 70B Q4_K_M & Qwen 3.8 27B Q8.
โ
Runs 100% in GPU VRAM (Optimal Headroom)21.93 GB (46%)~34 t/s
Qwen 3.8 27B Instruct on 2x NVIDIA RTX 3090 Dual Rig (48GB Total VRAM)
โ Fits Comfortably in VRAM
โ FITS IN VRAM
Required VRAM
21.93 GB
46% of hardware capacity (48 GB)Est. Decode Speed
~34 tok/s
Dense model @ 936 GB/sAvailable Headroom
+26.07 GB
Safe for OS display & full contextMemory Allocation & Headroom
21.93 GB allocated of 48 GB capacity
46% utilized
Weights: 16.32 GB
KV: (32K): 4.29 GB
Overhead: 1.32 GB
Free Headroom: +26.07 GB
โก Architecture & Throughput Transparency: Single-user decode speed (~34 tok/s) is memory-bandwidth bound at batch size = 1 on 2x NVIDIA RTX 3090 Dual Rig (48GB Total VRAM) (936 GB/s). Prompt prefill / TTFT (~493 tok/s) is compute-bound and utilizes GPU Tensor cores. Actual latency varies with FlashAttention-2 and runtime engine (Ollama/vLLM/llama.cpp).
OLLAMA_FLASH_ATTENTION=1 OLLAMA_KV_CACHE_TYPE=q8_0 ollama run qwen3.8:27bCloud Instance Options for Qwen 3.8 27B Instruct
On-DemandDeploy this model on an on-demand cloud GPU without local hardware constraints or power usage:
RunPod Cloud GPUsBest 1-Click Ollama & vLLM
$0.34/hr (RTX 4070 Ti) โ $2.39/hr (H100)Vast.ai GPU MarketplaceCheapest Dual RTX 3090 / 4090
$0.18/hr (RTX 3060) โ $0.55/hr (RTX 4090)Lambda GPU CloudEnterprise 8x H100 Clusters
$1.99/hr (GH200 / H100)Quantization Precision Matrix
32K ctx โข KV: Q8_0| Quantization | Weights | KV Cache | Total VRAM | Est. Speed | Fit Status |
|---|---|---|---|---|---|
| FP1616 bpw | 54.4 GB | 4.29 GB | 61.34 GB (128%) | ~2 t/s | PARTIAL |
| Q8_08.5 bpw | 28.83 GB | 4.29 GB | 34.88 GB (73%) | ~19 t/s | FITS |
| Q6_K6.56 bpw | 22.3 GB | 4.29 GB | 28.12 GB (59%) | ~25 t/s | FITS |
| Q5_K_M5.68 bpw | 19.31 GB | 4.29 GB | 25.03 GB (52%) | ~29 t/s | FITS |
| Q4_K_M4.8 bpw | 16.32 GB | 4.29 GB | 21.93 GB (46%) | ~34 t/s | FITS |
| FP44.25 bpw | 14.42 GB | 4.29 GB | 19.96 GB (42%) | ~38 t/s | FITS |
| IQ3_XS3.3 bpw | 11.15 GB | 4.29 GB | 16.58 GB (35%) | ~48 t/s | FITS |
Export Full Architecture & Infrastructure Spec
Get an executive-ready PDF report with exact memory math, hardware benchmarks, and cost comparisons.
All Open-Weights Models Tested on 2x NVIDIA RTX 3090 Dual Rig (48GB Total VRAM) (Q4_K_M @ 32K Context)
Standard Q4_K_M precision with 32K context window and Q8_0 KV cache
| Model | Architecture | Total VRAM Needed | Headroom | Est. Decode Speed | Compatibility | Detailed Config |
|---|---|---|---|---|---|---|
| Qwen 3.8 27B Instruct Alibaba Cloud | 27.2B Dense | 21.93 GB(46%) | +26.07 GB | ~34 tok/s | โ FITS | Configure CLI โ |
| Qwen 3.6 35B-A3B MoE Alibaba Cloud | 35B MoE | 24.1 GB(50%) | +23.9 GB | ~263 tok/s | โ FITS | Configure CLI โ |
| DeepSeek-V4.1-Flash (552B Engram MoE) DeepSeek | 552B MoE | 345.59 GB(720%) | -297.59 GB | 0 (OOM) | โ OOM | Configure CLI โ |
| Meta Muse Spark Open (816B-A42B MoE) Meta AI | 816B MoE | 512.86 GB(1068%) | -464.86 GB | 0 (OOM) | โ OOM | Configure CLI โ |
| Google Gemma 4 26B MoE Google | 26.4B MoE | 20.23 GB(42%) | +27.77 GB | ~176 tok/s | โ FITS | Configure CLI โ |
| Google Gemma 4 E4B (Edge 4B) Google | 4.1B Dense | 4.24 GB(9%) | +43.76 GB | ~215 tok/s | โ FITS | Configure CLI โ |
| DeepSeek-R1 Distill Qwen 32B DeepSeek | 32.8B Dense | 25.41 GB(53%) | +22.59 GB | ~28 tok/s | โ FITS | Configure CLI โ |
| DeepSeek-R1 Distill Qwen 14B DeepSeek | 14.8B Dense | 13.16 GB(27%) | +34.84 GB | ~60 tok/s | โ FITS | Configure CLI โ |
| Meta Llama 4 Scout (109B-A17B MoE) Meta AI | 109B MoE | 71.66 GB(149%) | -23.66 GB | 0 (OOM) | โ OOM | Configure CLI โ |
| Meta Llama 3.3 70B Instruct Meta AI | 70.6B Dense | 49.96 GB(104%) | -1.96 GB | ~3 tok/s | โก PARTIAL | Configure CLI โ |
| Meta Llama 3.1 8B Instruct Meta AI | 8B Dense | 7.87 GB(16%) | +40.13 GB | ~109 tok/s | โ FITS | Configure CLI โ |
| Microsoft Phi-4 Mini 3.8B Microsoft | 3.8B Dense | 3.92 GB(8%) | +44.08 GB | ~234 tok/s | โ FITS | Configure CLI โ |