modelrigs

Qwen3 235B-A22B on NVIDIA RTX 6000 Ada 48 GB

235.094B (22B active) on 48 GB at 960 GB/s. Hardware details · Model details

No calculated fit. Qwen3 235B-A22B does not fit under standard overhead assumptions at Q4_K_M: the model calculates 143.3 GB required against 44.2 GB usable.
CompatibilityDoes not fit
Calculated fit
No
Calculated Requires more than 97% of usable memory. Needs a smaller quantization or more memory.
Smallest catalogued quantization
Q4_K_M
Calculated gguf
Memory required
143.3 GB
Calculated of 44.2 GB usable — 324%
Max practical context
0
Calculated model supports 256K

Memory budget at 8K context

Model weights ⓘ
136.7 GB Calculated
KV cache ⓘ
1.5 GB Calculated
Runtime overhead ⓘ
5.1 GB Estimated
Total required
143.3 GB Calculated
Headroom ⓘ
-99.1 GB Calculated
No quantization in our catalogue fits this machine.
Discrete GPU: 48 GB of VRAM, of which we assume 92% is usable after driver and context overhead.
Mixture of experts: all 235.094B parameters must be resident in memory even though only ~22B are active per token. Memory follows total parameters; speed follows active parameters.
Performance

Estimated performance · recommended Q4_K_M

Decode, prefill and TTFT below are estimates for Q4_K_M. We do not have a comparable Q4_K_M measurement on this machine.

Estimated decode
—
Estimated Q4_K_M; calculated range
Estimated prefill
—
Estimated Q4_K_M; calculated range
Estimated TTFT at 8K
—
Estimated Q4_K_M; calculated range
Hardware load reference
300 W
Official spec hardware TDP; not this model run
No throughput figures for this configuration. We do not estimate performance for a model that cannot load.
How the estimate is calculated
  1. Decode: reading 22.0B active parameters at 4.85 bits/weight takes 19.85 ms at 960 GB/s x 70% achieved efficiency.
  2. MoE routing penalty of 15% applied: expert gathers are less bandwidth-efficient than a dense sweep.
  • MoE decode depends on how well the engine batches expert gathers; real results vary more than for dense models.
  • No FP16 compute figure for this hardware, so prefill cannot be estimated.
  • 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.

Submit a benchmarkSubmissions stay pending until reviewed.
This model at other quantizations on this machine
QuantizationResidentDiskTotal neededUtilisationFitMax context
Q4_K_Msmallest listed136.7 GB136.7 GB143.3 GB324%Does not fit0
Q5_K_M160.4 GB160.4 GB167.7 GB380%Does not fit0
Q8_0237.3 GB237.3 GB246.9 GB559%Does not fit0
BF16437.9 GB437.9 GB453.5 GB1027%Does not fit0
Cost of running Qwen3 235B-A22B on this machineUnited States (federal) · C corporation · 8h/day
Monthly economic cost
$2
Calculated after tax
Codex/Claude Code
$100/mo
≈ $100/mo · local is $98 less
Net cash at purchase
$0
Calculated VAT not reclaimable
Total over 5 years
$104
Calculated after tax, after resale
Cost per USD/1M tokens
—
Calculated

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
0 over 5 years, straight-line to a 0 residual
$0.00
Electricity
15.6 kWh/month at 0.140/kWh
$2.19
Monthly cost before tax$2.19
Electricity tax shield
Running costs are deductible business expenses
−$0.46
Monthly economic cost after tax$1.73

Three different numbers, deliberately

Cash cost
$0
Money that leaves the bank account on day one, net of reclaimable VAT.
Accounting depreciation
$0.00/month
$0 written down over 5 years to a $0 residual.
After-tax economic cost
$1.73/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
$0 (15%) Assumption
Two GPU generations later, the card is worth a fraction of its list price.
Electricity
$0.140/kWh Assumption
Average power draw
89 W Calculated
Load 300 W for 20% of powered hours, idle 36 W for the rest — a machine that is on is not generating tokens the whole time.
Investment allowances
§179 100.0% Official spec
Worth $0 in total — first-year expensing that replaces later tax depreciation.
Cost of capital
not applied Assumption
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.
Qwen3 235B-A22B on other hardware
HardwareDecodePriceFit
NVIDIA DGX H200 (8x H200, 1,128 GB)~341–490 t/s$359,430 *Comfortable
Dell Precision 7960 Rack (2x RTX PRO 6000, 192 GB)~81.3–117 t/s$36,663 *Fits
Lenovo ThinkStation PX (4x RTX PRO 6000, 384 GB)~70.3–101 t/s$58,553 *Comfortable
Mac Studio M5 Ultra 256 GB~44.3–63.8 t/s$8,647 *Comfortable
NVIDIA DGX Station GB300 (748 GB)~30.6–44 t/s$90,082 *Comfortable
Mac Studio M5 Ultra 512 GB~30.2–43.5 t/s$11,350 *Comfortable
Nearest alternatives to the NVIDIA RTX 6000 Ada 48 GB
Qwen3 235B-A22B on Dual used RTX 4090 workstation (48 GB)48 GBQwen3 235B-A22B on NVIDIA RTX A6000 48 GB48 GBQwen3 235B-A22B on NVIDIA A100 PCIe 40 GB40 GBQwen3 235B-A22B on Dual RTX 5090 workstation (2x 32 GB)64 GBQwen3 235B-A22B on NVIDIA RTX 5000 Ada 32 GB32 GB
Compare NVIDIA RTX 6000 Ada 48 GB against Dual used RTX 4090 workstation (48 GB) →