Appearance
Explore LLM and image benchmark results for NVIDIA DGX Spark, with hardware specifications and the software settings used in each test.
| Model | Benchmark | Configuration | DGX Spark | Olares One |
|---|---|---|---|---|
| Qwen3.8-27B | IFBench Evaluation time | llama.cpp, NVFP4 LOW | 18.2 h | 8.7 h |
| Qwen3.8-27B | Inference speed Token generation, single request | llama.cpp, NVFP4 LOW | 45 tok/s | 93 tok/s |
| Qwen3.8-Flash-Next | Inference speed Token generation, single request | Strata, Q2_0 | 80 tok/s | 166 tok/s |
| Qwen-Image-2.1 Base | Text to image Per image | ComfyUI, NVFP4 | 27 s | 12 s |
| Qwen-Image-2.1 Base | Image edit Per image | ComfyUI, NVFP4 | 37 s | 17 s |
| Qwen-Image-2.1 Turbo | Image edit Per image | ComfyUI, int8 | 13 s | 5.2 s |
llama.cpp, NVFP4 LOW
--mmproj mmproj-F16.gguf -c 104448 -ngl all -fa on -ctk q8_0 -ctv q8_0 --jinja -np 1 --spec-type draft-dflash -md Qwen3.8-27B-DFlash2-Q8_0.gguf -ngld all --spec-draft-n-max 7llama.cpp, NVFP4 LOW
--mmproj mmproj-F16.gguf -c 104448 -ngl all -fa on -ctk q8_0 -ctv q8_0 --jinja -np 4 -kvu --spec-type draft-dflash -md Qwen3.8-27B-DFlash2-Q8_0.gguf -ngld all --spec-draft-n-max 7Strata, Q2_0
--pack q2_0 --native Qwen3.8-Flash-Next-GSQ-RCO-Q2_0-00001-of-00002.gguf --ple-gguf Qwen3.8-Flash-Next-GSQ-RCO-Q2_0-00002-of-00002.gguf --expert-profile expert-profile.bin --expert-cache auto --prefill auto:32768 --spec 4 --spec-min-p 0.5 --mtp rt --max-context 131072 --kv int8 --mmap-experts --mtp-q4 all --mtp-hnorm stream --ple-io ram --kv-resident 32768 --adapt-every 100000Strata, IQ3_XXS
--pack iq3_xxs --native Qwen3.8-Flash-Next-GSQ-RCO-IQ3_XXS-00001-of-00002.gguf --ple-gguf Qwen3.8-Flash-Next-GSQ-RCO-IQ3_XXS-00002-of-00002.gguf --expert-profile expert-profile.bin --expert-cache auto --prefill auto:32768 --spec 4 --spec-min-p 0.5 --mtp rt --max-context 131072 --kv int8 --mmap-experts --mtp-q4 all --mtp-hnorm stream --ple-io ram --kv-resident 32768 --adapt-every 100000Strata, IQ3_S
--pack iq3_s --native Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S-00001-of-00002.gguf --ple-gguf Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S-00002-of-00002.gguf --expert-profile expert-profile.bin --expert-cache auto --prefill auto:32768 --spec 4 --spec-min-p 0.5 --mtp rt --max-context 131072 --kv int8 --mmap-experts --mtp-q4 all --mtp-hnorm stream --ple-io ram --kv-resident 32768 --adapt-every 100000Strata, UD-IQ4_XS
--pack unsloth-ud-iq4_xs --native Qwen3.8-Flash-Next-UD-IQ4_XS-00001-of-00003.gguf --expert-profile expert-profile.bin --expert-cache 16405 --prefill auto:32768 --spec 4 --spec-min-p 0.5 --mtp rt --max-context 131072 --kv int8 --mmap-experts --mtp-q4 all --mtp-hnorm stream --ple-io ram --kv-resident 32768 --resident-budget-gib 55 --adapt-every 100000Qwen-Image-2.1 Base: Text to image, Image edit
Qwen-Image-2.1 Turbo: Image edit
--listen 127.0.0.1 --port 8288 --extra-model-paths-config extra_model_paths.yaml --output-directory images --input-directory inputsQwen-Image-2.1 Base: Text to image, Image edit
Qwen-Image-2.1 Turbo: Image edit
--listen 127.0.0.1 --port 8288 --extra-model-paths-config extra_model_paths.yaml --output-directory images --input-directory inputs