Skip to main contentThe Local Bench

CPU only · Ryzen 9 9950X + 128 GB DDR5

AMD · CPU reference · as of 2026-10-12

Specs sourced · speeds estimated
Home server

  • 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

128GB
memory, 96.0 GB usableSourced
90GB/s
memory bandwidthSourced
$1,400
price, USDSample
10of 10
catalogue models fit at 8k

The no-GPU baseline: dual-channel DDR5 at 89.6 GB/s. Fine for MoE, slow for dense.

Decode-speed ceiling for a model is 90 GB/s ÷ GB read per token. For a 16.6 GB file that is 5 tok/s; real runtimes reach roughly 60% of it. Why bandwidth is the limit.

See every model on this device

Specifications

Memory
128 GBSourcedBuild configuration (2 × 64 GB); amd.com lists 256 GB max, 2 channels. Source
Bandwidth
89.6 GB/sSourced2 channels × DDR5-5600 (amd.com max supported speed for 2 DIMMs) × 8 bytes = 89.6 GB/s (derived from vendor figures). Source
FP16 dense compute
2.5 TFLOPSSampleSample FP16 dense compute; pending vendor-sheet citation.
Vendor TDP / board power
170 WSourcedDefault TDP, amd.com. Source
Load power at wall
200 WSampleSample wall power under load.
Price, USD
$1,400SampleSample price.
Price, CAD
CA$1,998SampleUSD price × 1.4271 (Bank of Canada USD/CAD, 2026-10-09); no Canadian retail listing captured. Converted from a sample USD value. Source

Models that run

Default quant, single stream, 8k context. Fastest first; estimates are ±25% until measured.

Model
Quant
Memory
Verdict
Decode, tok/s
Max context
gpt-oss-20bMXFP412.9 GBFits19.6Estimate128k
Qwen3 30B-A3BQ4_K_M19.9 GBFits15.9Estimate32k
gpt-oss-120bMXFP464.3 GBFits14.6Estimate128k
Llama 3.1 8B InstructQ4_K_M6.6 GBFits7.5Estimate128k
Qwen3 8BQ4_K_M6.8 GBFits7.2Estimate32k
Qwen3 14BQ4_K_M10.9 GBFits4.3Estimate32k
Mistral Small 3.2 24BQ4_K_M16.2 GBFits2.9Estimate128k
Gemma 3 27B ITQ4_K_M18.2 GBFits2.5Estimate128k
Qwen3 32BQ4_K_M22.5 GBFits2.0Estimate32k
Llama 3.3 70B InstructQ4_K_M45.8 GBFits1.0Estimate128k

Usable memory: discrete VRAM − 0.5 GB reserve; unified RAM × 0.75; fits when total ≤ 90% of usable (concept brief §7.3). Markdown · JSON