{"as_of":"2026-10-12","license":"CC-BY-4.0","license_url":"https://creativecommons.org/licenses/by/4.0/","cite_as":"The Local Bench (thelocalbench.com), data as of 2026-10-12, CC BY 4.0","source_url":"https://thelocalbench.com","hardware":{"id":"dgx-spark-128g","name":"NVIDIA DGX Spark FE 128 GB","vendor":"NVIDIA","class":"Unified memory","kind":"unified","as_of":"2026-10-12","summary":"CUDA on 128 GB of unified memory. Similar bandwidth to Strix Halo, more compute for prefill.","memory_gb":{"value":128,"flag":"sourced","unit":"GB","source":"https://www.nvidia.com/en-us/products/workstations/dgx-spark/","note":"Vendor spec page, checked 2026-10-12."},"bandwidth_gbs":{"value":273,"flag":"sourced","unit":"GB/s","source":"https://www.nvidia.com/en-us/products/workstations/dgx-spark/","note":"Vendor spec page, checked 2026-10-12."},"fp16_tflops":{"value":100,"flag":"sample","unit":"TFLOPS","note":"Sample FP16 dense compute; pending vendor-sheet citation."},"host_ram_gb":{"value":0,"flag":"sample","unit":"GB","note":"Assumed host RAM for offload."},"host_bandwidth_gbs":{"value":273,"flag":"sample","unit":"GB/s"},"load_watts":{"value":170,"flag":"sample","unit":"W","note":"Sample wall power under load."},"tdp_w":{"value":140,"flag":"sourced","unit":"W","source":"https://www.nvidia.com/en-us/products/workstations/dgx-spark/","note":"GB10 TDP (CPU+GPU), nvidia.com; system power supply is 240 W."},"price_usd":{"value":4699,"flag":"sourced","unit":"USD","source":"https://vucense.com/tech-reviews/compute-chips/dgx-spark-vs-ryzen-ai-max-395-local-ai-workstation-2026/","note":"US$4,699 (secondary source)."},"apple":false,"price_cad":{"value":6705.94,"flag":"estimate","unit":"CAD","source":"https://www.bankofcanada.ca/valet/observations/FXUSDCAD/json?recent=1","note":"USD price × 1.4271 (Bank of Canada USD/CAD, 2026-10-09); no Canadian retail listing captured."}},"runs":[{"model":"gemma-3-27b","hardware":"dgx-spark-128g","quant":"Q4_K_M","runtime":"llama.cpp (CUDA)","test":"tg128 @ 8k depth","tok_s":7.4,"min":7.2,"max":7.6,"repeats":5,"flag":"sample","date":"2026-10-12","note":"Placeholder figure from the approved mockup; not a measurement."},{"model":"llama-2-7b","hardware":"dgx-spark-128g","quant":"Q4_0","runtime":"llama.cpp (CUDA)","test":"tg128","tok_s":57.21,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/discussions/15013","note":"Llama 2 7B Q4_0, llama-bench tg128, full GPU offload, flash attention off, single GPU. Community scoreboard entry; commit and submitter in the source table.","min":57.15,"max":57.27},{"model":"gpt-oss-20b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 0","tok_s":83.43,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":82.84,"max":84.02,"depth":0},{"model":"gpt-oss-20b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 4096","tok_s":79.22,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":78.62,"max":79.82,"depth":4096},{"model":"gpt-oss-20b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 8192","tok_s":75.22,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":74.17,"max":76.27,"depth":8192},{"model":"gpt-oss-20b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 16384","tok_s":70.36,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":69.99,"max":70.73,"depth":16384},{"model":"gpt-oss-20b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 32768","tok_s":61.65,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":61.35,"max":61.95,"depth":32768},{"model":"gpt-oss-20b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ 8k depth","tok_s":75.22,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":74.17,"max":76.27},{"model":"gpt-oss-120b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 0","tok_s":58.72,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":58.52,"max":58.92,"depth":0},{"model":"gpt-oss-120b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 4096","tok_s":55.67,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":55.32,"max":56.02,"depth":4096},{"model":"gpt-oss-120b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 8192","tok_s":52.87,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":52.44,"max":53.3,"depth":8192},{"model":"gpt-oss-120b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 16384","tok_s":49.45,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":49.25,"max":49.65,"depth":16384},{"model":"gpt-oss-120b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ depth 32768","tok_s":42.76,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":42.62,"max":42.9,"depth":32768},{"model":"gpt-oss-120b","hardware":"dgx-spark-128g","quant":"MXFP4","runtime":"llama.cpp (CUDA)","test":"tg32 @ 8k depth","tok_s":52.87,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark.","min":52.44,"max":53.3},{"model":"qwen3-coder-30b-a3b-instruct","hardware":"dgx-spark-128g","quant":"Q8_0","runtime":"llama.cpp (CUDA)","test":"tg32 @ 8k depth","tok_s":48.02,"flag":"sourced","date":"2026-10-12","source":"https://github.com/ggml-org/llama.cpp/blob/master/benches/dgx-spark/dgx-spark.md","note":"llama-bench -fa 1 -p 2048 -n 32 -ub 2048, ggml-org GGUF, commit 3795cc1e8 (2026-02-05), run by the llama.cpp maintainers on a DGX Spark. Model is Qwen3-Coder-30B-A3B-Instruct. Catalogue correction 2026-10-13: model identifier corrected from Qwen3-30B-A3B; shared architecture does not establish identical performance.","min":47.62,"max":48.42}]}