Dual used RTX 4090 workstation (48 GB)
Discontinued but abundant on the used market, and still the best value per GB/s if you can live with 24 GB. The reference point for 'is a used build cheaper than a new appliance'.
NVIDIAcomponentAvailable16 models fit0 measured
Memory
48 GB
Official spec GDDR6X
Aggregate theoretical bandwidth
2,016 GB/s
Official spec raw sum; usable scaling is modelled separately
Specifications
CPU
AMD Ryzen 9 7950X Official spec
GPU
2× 2x GeForce RTX 4090 24 GB (used) Official spec
Memory
48 GB GDDR6X Official spec
Aggregate theoretical bandwidth
2,016 GB/s raw sum Official spec source
Multi-GPU scaling ⓘ
Interconnect-adjusted in estimator Assumption
FP16 compute ⓘ
~330 TFLOPS Official spec
System RAM
128 GB Official spec
Architecture
Discrete GPU with dedicated VRAM Official spec
Nominal power
1150 W Official spec
Measured load power
980 W Measured
Idle power
95 W Measured
Released
12 Oct 2022 Official spec
48 GB for roughly the price of a Mac Studio M5 Max 64 GB, at about eight times the power draw. The classic used-hardware trade: much better throughput per euro of purchase price, much worse throughput per euro of running cost.
Standout models on this machine
Fastest
gpt-oss-20b — ~238–343 t/s
MXFP4 · estimated confidence
Largest that fits
Related hardware
Model performance on Dual used RTX 4090 workstation (48 GB)0 measured, 16 estimated
16 of 16 rows
| Model↕ | Quant | Memory↕ | Decode▼ | Prefill↕ | Context | Fit | Confidence↕ |
|---|---|---|---|---|---|---|---|
| gpt-oss-20b OpenAI · 20.915B (3.6B active) | MXFP4recommended | 13.4 GB | ~238–343 t/s | ~12850–26700 t/s | 128K | Comfortable | Estimated |
| Nemotron 3.5 Lightning 30B-A3B NVIDIA · 31.578B (3B active) | Q5_K_Mrecommended | 23.7 GB | ~223–321 t/s | ~5730–11900 t/s | 1024K | Comfortable | Estimated |
| Qwen3-Coder 30B-A3B Alibaba Qwen · 30.532B (3.3B active) | Q5_K_Mrecommended | 23.6 GB | ~210–302 t/s | ~5560–11540 t/s | 192K | Comfortable | Estimated |
| Qwen3 30B-A3B Alibaba Qwen · 30.532B (3.3B active) | Q5_K_Mrecommended | 23.6 GB | ~210–302 t/s | ~5560–11540 t/s | 40K | Comfortable | Estimated |
| Gemma 4 26B-A4B Google DeepMind · 25.806B (3.8B active) | Q8_0recommended | 30.1 GB | ~143–206 t/s | ~5630–11700 t/s | 32K | Comfortable | Estimated |
| Qwen3 8B Alibaba Qwen · 8.191B | Q8_0recommended | 11.0 GB | ~88.1–127 t/s | ~6810–14140 t/s | 40K | Comfortable | Estimated |
| Phi-4 14B Microsoft · 14.66B | Q8_0recommended | 18.2 GB | ~52.7–75.9 t/s | ~3800–7900 t/s | 16K | Comfortable | Estimated |
| Gemma 3 27B Google DeepMind · 27.432B | Q5_K_Mrecommended | 24.6 GB | ~42.9–61.7 t/s | ~2030–4220 t/s | 32K | Comfortable | Estimated |
| Granite 4.2 30B IBM · 29.277B | Q5_K_Mrecommended | 24.0 GB | ~40.4–58.1 t/s | ~1900–3960 t/s | 64K | Comfortable | Estimated |
| Gemma 4 31B Google DeepMind · 31.273B | Q5_K_Mrecommended | 30.9 GB | ~38–54.6 t/s | ~1780–3700 t/s | 16K | Comfortable | Estimated |
| Qwen3 32B Alibaba Qwen · 32.762B | Q5_K_Mrecommended | 26.4 GB | ~36.3–52.3 t/s | ~1700–3540 t/s | 40K | Comfortable | Estimated |
| DeepSeek-R1-Distill 32B DeepSeek · 32.764B | Q5_K_Mrecommended | 26.4 GB | ~36.3–52.3 t/s | ~1700–3540 t/s | 64K | Comfortable | Estimated |
| Devstral Small 24B Mistral AI · 23.572B | Q8_0recommended | 27.2 GB | ~33.9–48.9 t/s | ~2370–4910 t/s | 96K | Comfortable | Estimated |
| Mistral Small 3.2 24B Mistral AI · 24.011B | Q8_0recommended | 27.6 GB | ~33.4–48 t/s | ~2320–4820 t/s | 64K | Comfortable | Estimated |
| Qwen3.8 27B Alibaba Qwen · 27.781B | Q8_0recommended | 30.8 GB | ~29.1–41.8 t/s | ~2010–4170 t/s | 128K | Comfortable | Estimated |
| Qwen3.6 27B Alibaba Qwen · 27.781B | Q8_0recommended | 30.8 GB | ~29.1–41.8 t/s | ~2010–4170 t/s | 128K | Comfortable | Estimated |
Rows in grey are estimates from our bandwidth model, not measurements — they are shown as a range and never as a precise figure. Use the “measured only” filter to see just the 0 pairings on this machine that a real benchmark backs.
Submit a benchmarkContributions are reviewed before publication.
Ownership economicsUnited States (federal) · C corporation · 8h/day
Monthly economic cost
$48
Calculated after tax
Codex/Claude Code
$100/mo
≈ $100/mo · local is $52 less
Net cash at purchase
$3,692
Calculated VAT not reclaimable
Total over 5 years
$2,873
Calculated after tax, after resale
Cost per USD/1M tokens
$1.30
Calculated 36.8M tokens/month
The $100 comparison uses the Codex Pro 5x / Claude Max 5x plans. A fixed planning conversion is used for USD. This compares monthly spend only: subscriptions have usage limits, local hardware has different capabilities and constraints, and taxes or regional pricing may change the charged amount. Prices checked 7 September 2026.
Purchase-price input: $3,692 in United States (federal), tax/VAT excluded · estimated · Used market (Marktplaats/eBay) (converted catalogue estimate). This is the localized purchase input; the after-tax economic cost is calculated separately below.
Monthly breakdown
| Depreciation 2,880 over 5 years, straight-line to a 812 residual | $48.00 |
| Electricity 47.9 kWh/month at 0.140/kWh | $6.70 |
| Cost of capital 4.0%/yr on 2,252 average capital employed | $7.51 |
| Monthly cost before tax | $62.21 |
| Electricity tax shield Running costs are deductible business expenses | −$1.41 |
| First-year expensing §179 (100.0%) | −$12.92 |
| Monthly economic cost after tax | $47.88 |
Three different numbers, deliberately
Cash cost
$3,692
Money that leaves the bank account on day one, net of reclaimable VAT.
Accounting depreciation
$48.00/month
$2,880 written down over 5 years to a $812 residual.
After-tax economic cost
$47.88/month
Depreciation plus running costs plus cost of capital, less the tax those deductions save. This is the figure to compare between machines.
Assumptions and sources (verified 2026-09-07)
VAT rate
0% Official spec
VAT recoverable
0% Official spec
Federal C-corporation estimate at the flat 21% rate. Pass-through entities should use the sole-proprietor estimate as a rougher proxy.
Effective deduction rate
21.00% Calculated
Headline marginal rate 21.00%.
Depreciation
5 years, straight-line Assumption
Residual value
$812 (22%) Assumption
Extrapolated.
Electricity
$0.140/kWh Assumption
Average power draw
272 W Calculated
Load 980 W for 20% of powered hours, idle 95 W for the rest — a machine that is on is not generating tokens the whole time.
Investment allowances
§179 100.0% Official spec
Worth $775 in total — first-year expensing that replaces later tax depreciation.
Cost of capital
$7.51/month Assumption
- VAT rate: No US federal VAT · verified 2026-09-07
- Marginal tax rate: IRS Publication 542 — Corporations · verified 2026-09-07
- VAT recoverable fraction: site assumption · verified 2026-09-07
- Useful life: IRS Publication 946 — How To Depreciate Property · verified 2026-09-07
- Electricity price: Site assumption · verified 2026-09-07
Assumes the machine generates tokens 20% of its 176 powered hours per month. Cost per token scales inversely with this number — halve the utilisation and the cost per token doubles.
Federal planning estimate only, not tax advice. State and local income tax, sales/use tax and incentives are excluded. It assumes 100% business use, a Section 179 election, enough business income to use it, and that the full annual limit remains available.