Gemma 4 31B on Dual RTX 5090 workstation (2x 32 GB)
31.273B on 64 GB at 3,584 GB/s. Hardware details · Model details
CompatibilityComfortable
Fits
Yes
Calculated Uses under 80% of usable memory. Room for a long context and other work.
Recommended quantization
Q8_0
Calculated gguf
Memory required
41.4 GB
Calculated of 56.3 GB usable — 74%
Max practical context
16K
Calculated model supports 256K
Memory budget at 8K context
Model weights ⓘ
31.6 GB Calculated
KV cache ⓘ
7.5 GB Calculated
Runtime overhead ⓘ
2.3 GB Estimated
Total required
41.4 GB Calculated
Usable memory ⓘ
56.3 GB Assumption
Headroom
14.9 GB Calculated
Highest-precision quantization that leaves headroom: uses 74% of usable memory at 8K context.
2 GPUs pooling 64 GB total VRAM. Assumes tensor- or layer-parallel sharding; each GPU carries its own context and communication buffers.
Multi-GPU throughput depends heavily on interconnect. Without NVLink, tensor parallelism over PCIe adds latency that partially offsets the extra bandwidth.
PerformanceEstimated0/10
Decode (generation)
~45.2–65.1 t/s
Estimated calculated, not measured
Prefill (prompt)
~1600–3330 t/s
Estimated
TTFT at 8K
~2.5–5.2 s
Estimated time to first token
Power while generating
1180 W
Measured 0.17 tokens/s per watt
These figures are estimated, not measured. Nobody has published a benchmark of Gemma 4 31B on this machine that we know of, so we calculate throughput from memory bandwidth and active parameter count. Estimates are always shown as a range. Expand the working below to see exactly how the number was produced.
How the estimate is calculated
- Multi-GPU: 1792 GB/s per card, with each additional card contributing 40% of its bandwidth — 2509 GB/s effective, not the 3584 GB/s aggregate. Consumer Blackwell has no NVLink, so every layer's all-reduce crosses PCIe.
- Decode: reading 31.3B active parameters at 8.50 bits/weight takes 16.73 ms at 2509 GB/s x 79% achieved efficiency.
- Prefill: 419 TFLOPS (FP16) x 2 for native FP8 tensor cores x 0.71 calibrated against measured prefill on this platform x 26% assumed model-FLOPs utilisation, divided by 2 x 31.3B parameters per token.
- Prefill throughput is highly engine-dependent. Flash attention, batch size and quantized KV all move this number substantially, and for sparse mixture-of-experts models it is the least reliable figure we produce.
- This is a calculated estimate, not a measurement. It assumes a single request, a short prompt, no speculative decoding and a warm model already resident in memory.
Estimator version estimator_v1. Stored with every estimated row so old estimates can be regenerated when the model improves.
This model at other quantizations on this machine
| Quantization | Weights | Total needed | Utilisation | Fit | Max context |
|---|---|---|---|---|---|
MLX 4-bit | 17.0 GB | 26.4 GB | 47% | Comfortable | 32K |
Q4_K_M | 18.2 GB | 27.6 GB | 49% | Comfortable | 32K |
Q8_0recommended | 31.6 GB | 41.4 GB | 74% | Comfortable | 16K |
BF16 | 58.3 GB | 68.9 GB | 122% | Does not fit | 0 |
Cost of running Gemma 4 31B on this machineUnited States (federal) · C corporation · 8h/day
Monthly economic cost
$80
Calculated after tax
Codex/Claude Code
$100/mo
≈ $100/mo · local is $20 less
Net cash at purchase
$5,854
Calculated VAT not reclaimable
Total over 5 years
$4,799
Calculated after tax, after resale
Cost per USD/1M tokens
$11.44
Calculated 7.0M 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.
Monthly breakdown
| Depreciation 4,976 over 5 years, straight-line to a 878 residual | $82.94 |
| Electricity 57.0 kWh/month at 0.140/kWh | $7.98 |
| Cost of capital 4.0%/yr on 3,366 average capital employed | $11.22 |
| Monthly cost before tax | $102.14 |
| Electricity tax shield Running costs are deductible business expenses | −$1.68 |
| First-year expensing §179 (100.0%) | −$20.49 |
| Monthly economic cost after tax | $79.98 |
Three different numbers, deliberately
Cash cost
$5,854
Money that leaves the bank account on day one, net of reclaimable VAT.
Accounting depreciation
$82.94/month
$4,976 written down over 5 years to a $878 residual.
After-tax economic cost
$79.98/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
$878 (15%) Assumption
Two GPU generations later, the card is worth a fraction of its list price.
Electricity
$0.140/kWh Assumption
Average power draw
324 W Calculated
Load 1180 W for 20% of powered hours, idle 110 W for the rest — a machine that is on is not generating tokens the whole time.
Investment allowances
§179 100.0% Official spec
Worth $1,229 in total — first-year expensing that replaces later tax depreciation.
Cost of capital
$11.22/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.
Efficiency metrics
Decode per $1,000 spent
9.4 t/s
Calculated purchase price only
Decode per $100/month
69.0 t/s
Calculated after-tax ownership cost
Tokens per joule
0.17
Calculated same as tokens/s per watt
USD per 1M output tokens
$11.44
Calculated at 20% utilisation
Local versus hosted APIs
| Hosted model | USD/1M output | Break-even | API at your volume | Verdict | Comparison type |
|---|---|---|---|---|---|
| Qwen3.8 Flash Alibaba Cloud | $0.42 | 51.9M/mo | $11 | API cheaper | Different model |
| DeepSeek-V4 Flash DeepSeek | $0.66 | 33.0M/mo | $17 | API cheaper | Different model |
| DeepSeek-V4 Pro DeepSeek | $1.98 | 11.0M/mo | $51 | API cheaper | Different model |
| DeepSeek V4 Pro (Together) Together AI | $4.40 | 3.8M/mo | $148 | Local cheaper | Different model |
| GLM-5.3 Z.ai | $4.40 | 5.1M/mo | $109 | Local cheaper | Different model |
| Claude Haiku 4.5 Anthropic | $5.00 | 6.2M/mo | $91 | Local cheaper | Different model |
Different-model comparison: the hosted model is not the model you would run locally. Treat this as a workload-quality trade-off, not a direct economic equivalence — the frontier model may complete a task in fewer tokens, or complete tasks the local model cannot. Break-even is the monthly output volume at which API spend equals the $80/month economic cost of owning this machine, assuming 8 input tokens per output token. Hosted prices are published in USD. This table is the canonical US default.
Gemma 4 31B on other hardware
| Hardware | Decode | Price | Fit |
|---|---|---|---|
| Quad RTX 5090 workstation (4x 32 GB) | ~68.1–98 t/s | €13.999 | Comfortable |
| RTX 5090 workstation (1x 32 GB) | ~62.1–89.4 t/s | €3.699 | Fits |
| Dual used RTX 4090 workstation (48 GB) | ~44–63.4 t/s | €4.099 | Comfortable |
| RTX PRO 6000 Blackwell workstation (96 GB) | ~31.8–45.8 t/s | €11.499 | Comfortable |
| RTX PRO 6000 Max-Q workstation (96 GB, 300 W) | ~31.8–45.8 t/s | €11.999 | Comfortable |
| Radeon AI PRO R9700 workstation (32 GB) | ~27.7–39.9 t/s | €2.799 | Fits |
Nearest alternatives to the Dual RTX 5090 workstation (2x 32 GB)
Gemma 4 31B on RTX 5090 workstation (1x 32 GB)32 GBGemma 4 31B on Quad RTX 5090 workstation (4x 32 GB)128 GBGemma 4 31B on Dual used RTX 4090 workstation (48 GB)48 GBGemma 4 31B on RTX PRO 6000 Blackwell workstation (96 GB)96 GBGemma 4 31B on RTX PRO 6000 Max-Q workstation (96 GB, 300 W)96 GB
Compare Dual RTX 5090 workstation (2x 32 GB) against RTX 5090 workstation (1x 32 GB) →
Compare Dual RTX 5090 workstation (2x 32 GB) against RTX 5090 workstation (1x 32 GB) →
Page quality score (why this page is or is not indexed)
10/13 — indexed. Generated pages are gated so we do not ask a search engine to rank a page with nothing computed to say. The directive is emitted in the page head via the metadata API, not in the body, so it is authoritative.
- Has a memory-fit calculation (+3)
- Has a performance estimate (+2)
- Has a price (+2)
- Has an ownership economics calculation (+2)
- Well connected (18 internal links) (+1)