Illustrative build · sample prices
Dual-GPU 48 GB
Two used 3090s split a 70B model by layers. Capacity doubles; single-stream bandwidth does not.
Data checked: 2026-10-12. Review interval shown below. Prices need review after 30 days; specifications and benchmark collections after 90 days. Dates describe the recorded check, not a live retailer quote.
- Verified Checked against the primary source (for example the model’s config.json on Hugging Face).
- Sourced Taken from a cited source; not yet independently checked.
- Estimate Calculated with the published formulas; not a measurement.
- Sample Placeholder data for layout; not a real figure.
- Not yet measured No public, citable score found for this benchmark yet. We never fill the gap with a guess.
Verified metadata is not a measured performance result. Only cited runs matching model, device, quant and recorded context can replace speed estimates. Estimates assume one stream and F16 KV cache, with a ±25% uncertainty band. A published run applies only to its recorded runtime and conditions; missing settings are unknown.
Evidence labels · Freshness policy · Calculation methodology
Includes the listed system parts. Excludes tax, shipping, peripherals, maintenance, financing and resale. Sample parts prices may differ from current hardware catalogue prices. Electricity assumes 780 W whole-system load for 4 h and 110 W idle for 20 h daily, 30 days at $0.15/kWh. Power is unmeasured.
Parts list
Sample parts and prices for layout.
Part | Item | Price, USD |
|---|---|---|
| GPU | 2× GeForce RTX 3090 24 GB (used) | $2,400Sample |
| CPU | Ryzen 9 7900 | $349Sample |
| Motherboard | X670E ATX (x8/x8) | $299Sample |
| Memory | 128 GB DDR5-5600 (4×32) | $359Sample |
| Storage | 4 TB NVMe Gen4 | $229Sample |
| PSU | 1 | $249Sample |
| Case | Full tower | $159Sample |
No affiliate links yet. Prices will be timestamped per retailer once tracking starts.
Expected speed
Reference models, default quant, 8k context.
Model | Verdict | tok/s |
|---|---|---|
| Llama 3.3 70B Instruct | Tight | 12.4Estimate |
| Qwen3 32B | Fits | 25.6Estimate |
| Gemma 3 27B IT | Fits | 31.8Estimate |
Break-even card
Electricity $23.94/month (780 W × 4 h + 110 W idle × 20 h, $0.15/kWh).
Plan | Price / month | Months to break even |
|---|---|---|
| ChatGPT Plus | $20 | never |
| Claude Pro | $20 | never |
| Google AI Pro | $19.99 | never |
| Claude Max (5×) | $100 | 53.2 |
| ChatGPT Pro 100 | $100 | 53.2 |
| Google AI Ultra (5×) | $99.99 | 53.2 |
| Claude Max (20×) | $200 | 23.0 |
| ChatGPT Pro 200 | $200 | 23.0 |
| Google AI Ultra (20×) | $199.99 | 23.0 |
| ChatGPT Pro 500 (top tier) | $500 | 8.5 |
Break-even = hardware ÷ (plan − electricity). It ignores capability: a local model is not a frontier model. Adjust the numbers.