DiggTop ⌕

DGX Spark vs RTX 4090 D

Pick a different pair →

Best sourced output tok/s per model (single-stream generation) — engine and version, quantization, context and a source link on every row. Hover a value for the quant.

DGX Spark52 sourced runs · best 86.8 tok/s
RTX 4090 D2 verified runs · best 70.9 tok/s

Specs and evidence

DGX SparkRTX 4090 D
VRAM128 GB24 GB
Sourced runs on the board522
Models measured91
Best output tok/s86.8 tok/s70.9 tok/s
Highest measured peak VRAM—23.5 GB
Street price (latest snapshot)——

Sourced rows only. Peak VRAM is recorded by the person who ran the benchmark, so it is a floor for what the card can hold, not a specification. Price snapshots.

Head-to-head

Only one model has been measured on both cards (Qwen3.8-27B). On that single data point DGX Spark records 56.0 tok/s and RTX 4090 D records 70.9 tok/s — treat it as one data point, not a ranking.

The measured matrix

ModelDGX SparkRTX 4090 D
Qwen 3.8 Flash86.8 EXL3 3.05bpw—
Qwen3.8-Flash-Next75.7 4.05bpw_h6_ng6—
Qwen3.8-27B56.0 NVFP4 W4A470.9 Q4_K_M
GLM-5.3-Flash64.1 EXL3 2.05 bpw—
CYBER-FROST-3.860.7 EXL3—
Qwen3.8-Flash-Next-hibrid4860.0 NVFP4 (output head)—
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B48.0 NVFP4—
DeepSeek-V4.1-Flash47.0 FP4—
Underdog-Saluki-27B22.6 IQ2-mix—

Head-to-head (models tested on both): DGX Spark 0 · RTX 4090 D 1 · tied 0 · data as of Oct 30, 2026. Full rules: methodology.

Related reading: RTX 5090 vs 3090 · quantization · Value index · Full board

↑