{ "id": "deepseek-1017-tensorfold-spark", "title": "DeepSeek 1017: TensorFold on two DGX Sparks", "kind": "benchmark", "source_type": "community_operator_report", "summary": "A traceable 1017 EXL3 deployment and two named decode configurations, with a separate later prefill cohort and explicit correctness limits.", "runtime": "TensorFold / vLLM", "runtime_topics": [ "TensorFold", "vLLM" ], "hardware_topics": [ "spark", "nvidia" ], "platform_topics": [ "linux" ], "model_label": "DeepSeek-V4.1-Flash-1017", "model_family": "DeepSeek", "model_id": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_release_date": null, "model_artifact": { "id": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1" }, "evidence_state": "source_reported", "site_reproduced": false, "record_type": "curated_reference", "performance_ranking_allowed": false, "metrics": null, "metrics_status": "site_metrics_unmeasured; attributed_source_figures_in_editorial.reported_results", "conditions": { "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "weight_hash": null, "runtime_execution_revision": null, "input_tokens": null, "output_tokens": null, "concurrency": null, "cache_state": null, "quantization": null, "speculative_decoding": null, "temperature": null, "top_p": null }, "reported_setup": { "system": "2 x DGX Spark GB10 TP2 over ConnectX-7", "node_count": 2, "parallelism": "TP2", "kv_dtype": null, "context_limit_tokens": null, "experiment_date": "2026-10-17 and 2026-10-18, separate cohorts" }, "test_profiles": [ { "label": "vllm-kit-run01 prose", "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/01_vllm-miaai-kit_suite-results_32of32.json#L6", "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "vllm-kit-run01", "engine": "vLLM + exllamav3 overlay", "engine_version": null, "runtime_execution_revision": null, "image_digest": "sha256:2f0cf3adc0f989c1d446be274df864eb799630175f604c3b22b71b7205971dce", "kit_revision": "6f7d1590ad49a2b8995188e45d7b9db31e677452", "draft_tokens": 3, "configured_max_requests": 1, "context_limit_tokens": 262144, "prefill_chunk_tokens": 1536, "chat_template": "kit template; actual CHAT_TEMPLATE override not read by the source", "workload": "prose", "per_prompt_decode_rates": [ 24.7, 29.3 ], "per_prompt_TTFT_seconds": [ 0.32, 0.28 ], "prompt_count": 2 } }, { "label": "vllm-kit-run01 code", "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/01_vllm-miaai-kit_suite-results_32of32.json#L20", "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "vllm-kit-run01", "engine": "vLLM + exllamav3 overlay", "engine_version": null, "runtime_execution_revision": null, "image_digest": "sha256:2f0cf3adc0f989c1d446be274df864eb799630175f604c3b22b71b7205971dce", "kit_revision": "6f7d1590ad49a2b8995188e45d7b9db31e677452", "draft_tokens": 3, "configured_max_requests": 1, "context_limit_tokens": 262144, "prefill_chunk_tokens": 1536, "chat_template": "kit template; actual CHAT_TEMPLATE override not read by the source", "workload": "code", "per_prompt_decode_rates": [ 40.4, 39.4 ], "per_prompt_TTFT_seconds": [ 0.47, 0.32 ], "prompt_count": 2 } }, { "label": "tensorfold-run11 prose", "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/11_tensorfold-fd2365042f56_suite-results_32of32.json#L6", "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "tensorfold-run11", "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": "fd2365042f56fcf726cf4071132c8b779ad2eae0", "image_tag": "tensorfold-dsv41:fd2365042f56", "image_digest": null, "draft_tokens": 5, "configured_max_requests": 4, "context_limit_tokens": 1048576, "prefill_chunk_tokens": 2048, "chat_template": "patched kit template; source infers per-run use from file times", "workload": "prose", "per_prompt_decode_rates": [ 53.4, 60.4 ], "per_prompt_TTFT_seconds": [ 0.12, 0.12 ], "prompt_count": 2 } }, { "label": "tensorfold-run11 code", "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/11_tensorfold-fd2365042f56_suite-results_32of32.json#L20", "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "tensorfold-run11", "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": "fd2365042f56fcf726cf4071132c8b779ad2eae0", "image_tag": "tensorfold-dsv41:fd2365042f56", "image_digest": null, "draft_tokens": 5, "configured_max_requests": 4, "context_limit_tokens": 1048576, "prefill_chunk_tokens": 2048, "chat_template": "patched kit template; source infers per-run use from file times", "workload": "code", "per_prompt_decode_rates": [ 91.1, 92.5 ], "per_prompt_TTFT_seconds": [ 0.13, 0.16 ], "prompt_count": 2 } }, { "label": "TensorFold final setup 32K prefill", "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-18/INDEX.md#L16", "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-18", "concurrency": 1, "thinking": false, "temperature": null, "top_p": null, "max_output_tokens": null, "output_tokens": null, "input_tokens": null, "prefix_cache": false, "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "prompt_tokens / TTFT; source final setup and host tuning", "repetitions": 5, "checkpoint_is_latest_revision_verified": false, "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": null, "runtime_revision_prefix": "92aa710b533d", "image_tag": "tensorfold-dsv41:92aa710b533d", "image_digest": null, "reported_prompt_bucket": "32K", "cached_tokens": 0, "per_run_values": [ 2201, 2180, 2065, 2165, 2100 ], "run_ids": [ "hw5", "hw6", "ht1", "ht2", "ht3" ], "draft_tokens": 5, "context_limit_tokens": 1048576, "warmup": "one 8K prompt; subsequent 8K requests before 32K/64K/128K; random-word prompts", "host_tuning_scope": "author applies no engine swap and memory-reclaim/compaction controls; not a universal host-tuning recommendation" } }, { "label": "TensorFold final setup 128K prefill", "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-18/INDEX.md#L16", "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-18", "concurrency": 1, "thinking": false, "temperature": null, "top_p": null, "max_output_tokens": null, "output_tokens": null, "input_tokens": null, "prefix_cache": false, "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "prompt_tokens / TTFT; source final setup and host tuning", "repetitions": 5, "checkpoint_is_latest_revision_verified": false, "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": null, "runtime_revision_prefix": "92aa710b533d", "image_tag": "tensorfold-dsv41:92aa710b533d", "image_digest": null, "reported_prompt_bucket": "128K", "cached_tokens": 0, "per_run_values": [ 1986, 1981, 1914, 1961, 1851 ], "run_ids": [ "hw5", "hw6", "ht1", "ht2", "ht3" ], "draft_tokens": 5, "context_limit_tokens": 1048576, "warmup": "one 8K prompt; subsequent 8K requests before 32K/64K/128K; random-word prompts", "host_tuning_scope": "author applies no engine swap and memory-reclaim/compaction controls; not a universal host-tuning recommendation" } } ], "deployment_options": [], "recipe_parameters": null, "measurement_scope": "inference_speed_and_task_checks", "notes": [ "The selected 1017 results describe two complete serving configurations, not an isolated engine-only test. Keep the converted weights and separate Engram revision, distinguish October 17 decode from October 18 prefill, and retain missing image, cache and quality evidence." ], "limitations": [ "Source-reported only; no independent site model runs.", "Document revision identifies the report, not the executed runtime or weight revision.", "Complete measured configurations and missing fields must accompany extracted values." ], "comparison_status": "not_comparable", "editorial_status": "source_derived_detailed_article", "public_benchmark_snapshot": false, "source": { "publisher": "Elevatormusic", "url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "repository": "Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "platform": "github", "revision": "9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32", "document_path": "README.md", "pinned_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/README.md", "report_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/README.md", "document_sha256": "546a0c3bfbbb6e84226ec812cf38927f6d993e53b73633a540c426eea5f13d29", "last_commit_at": "2026-10-18T05:40:29Z", "checked_on": "2026-10-18", "metadata_status": "checked" }, "supplemental_sources": [ { "document_id": "ds1017-docs-BENCHMARKS.md", "parent_record_id": "deepseek-1017-tensorfold-spark", "publisher": "Elevatormusic", "repository": "Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "revision": "9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32", "document_path": "docs/BENCHMARKS.md", "pinned_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/docs/BENCHMARKS.md", "document_sha256": "b2d05a077ae8145a2cf2a98ac2ae2eabef87b992230c32398a60e0a2e9174861", "checked_on": "2026-10-18" }, { "document_id": "ds1017-docs-KNOWN-LIMITS.md", "parent_record_id": "deepseek-1017-tensorfold-spark", "publisher": "Elevatormusic", "repository": "Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "revision": "9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32", "document_path": "docs/KNOWN-LIMITS.md", "pinned_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/docs/KNOWN-LIMITS.md", "document_sha256": "eeb6bc6f7a3fce20654f977581e9718b124e70fc01ab358efebefe9afd9dc942", "checked_on": "2026-10-18" }, { "document_id": "ds1017-runs-2026-10-17-PROVENANCE.md", "parent_record_id": "deepseek-1017-tensorfold-spark", "publisher": "Elevatormusic", "repository": "Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "revision": "9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32", "document_path": "runs/2026-10-17/PROVENANCE.md", "pinned_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/PROVENANCE.md", "document_sha256": "aa020bd6aa0fe840c9518c96dbc00f04bbdf6add834f2fe7bac2b076aa91a089", "checked_on": "2026-10-18" }, { "document_id": "ds1017-runs-2026-10-18-INDEX.md", "parent_record_id": "deepseek-1017-tensorfold-spark", "publisher": "Elevatormusic", "repository": "Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "revision": "9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32", "document_path": "runs/2026-10-18/INDEX.md", "pinned_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-18/INDEX.md", "document_sha256": "e69d8c270e6407ddbaacb82da36cc48994b6f20fd4300d176753371b4e23fb8e", "checked_on": "2026-10-18" }, { "document_id": "ds1017-runs-2026-10-17-01_vllm-miaai-kit_suite-results_32of32.json", "parent_record_id": "deepseek-1017-tensorfold-spark", "publisher": "Elevatormusic", "repository": "Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "revision": "9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32", "document_path": "runs/2026-10-17/01_vllm-miaai-kit_suite-results_32of32.json", "pinned_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/01_vllm-miaai-kit_suite-results_32of32.json", "document_sha256": "0cb7c61b5b321ba0b9eec76460e4d9d3d0e759bd09cfb03fc747ef66e7c2ea63", "checked_on": "2026-10-18" }, { "document_id": "ds1017-runs-2026-10-17-11_tensorfold-fd2365042f56_suite-results_32of32.json", "parent_record_id": "deepseek-1017-tensorfold-spark", "publisher": "Elevatormusic", "repository": "Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark", "revision": "9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32", "document_path": "runs/2026-10-17/11_tensorfold-fd2365042f56_suite-results_32of32.json", "pinned_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/11_tensorfold-fd2365042f56_suite-results_32of32.json", "document_sha256": "d0ec8c99b0208ac4ddc272379d3cd71c796b5662d09481d6ba1e98fbbf0fa41c", "checked_on": "2026-10-18" }, { "document_id": "official-deepseek-ai--DeepSeek-V4.1-Flash-1017", "parent_record_id": "deepseek-1017-tensorfold-spark", "publisher": "deepseek-ai", "repository": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "document_path": "README.md", "pinned_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017/blob/1bae0e1088993b9b82de08f59de5be5a34e42170/README.md", "document_sha256": "b72b796e4f4c739eca5d0ad360a7679b39118ca7eb64891f7151a527b292c42a", "checked_on": "2026-10-18" } ], "license": { "declared_spdx": "Apache-2.0", "scope": "Repository metadata only; model conversion, base weights, plugin and engine terms retain their own scope." }, "editorial": { "version": 1, "published_on": "2026-10-18", "modified_on": "2026-10-18", "byline": "Run AI Here editorial", "creation_method": "Local Spark Qwen draft, source-audited and edited by Run AI Here with Codex; no independent model execution.", "answer": "The selected 1017 results describe two complete serving configurations, not an isolated engine-only test. Keep the converted weights and separate Engram revision, distinguish October 17 decode from October 18 prefill, and retain missing image, cache and quality evidence.", "sections": [ { "heading": "Which 1017 artifact was measured?", "paragraphs": [ "The report uses an EXL3 2.9bpw conversion at afc7aa0aab98ee735c0f54e0a122019af1ac08f1 plus separate official 1017 Engram tables at 1bae0e1088993b9b82de08f59de5be5a34e42170. The author does not establish the measured conversion is the newest revision. This is a particular converted artifact, not a speed result for unconverted official weights." ], "citations": [ "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/PROVENANCE.md" ] }, { "heading": "What changes between the two decode arms?", "paragraphs": [ "Both use one request on the same reported model view and Spark pair. The vLLM kit has three drafts and a 262144 context cap; TensorFold run 11 has five drafts and a 1048576 cap. Their batching, configured concurrency and template paths also differ. The selected two-prompt means are observations of complete configurations, not an isolated engine effect." ], "citations": [ "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/docs/BENCHMARKS.md" ] }, { "heading": "Why not use the headline median as six repeats?", "paragraphs": [ "The report explicitly says its six TensorFold suite runs used different builds. Run 11 identifies fd2365042f56, whereas the current recipe builds 92aa710b533d. Keep those versions apart. The baseline has only one run, and there is no measured multi-user vLLM baseline." ], "citations": [ "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/docs/BENCHMARKS.md" ] }, { "heading": "What does the later prefill cohort establish?", "paragraphs": [ "Five October 18 source rows use the final setup, random-word prompts and recorded zero cached tokens. The selected 32K and 128K medians retain all five row values and their observed ranges. These are separate from the older repeated-sentence suite, where cache hits could not be excluded; range is not a confidence interval." ], "citations": [ "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-18/INDEX.md" ] }, { "heading": "Where are the limits?", "paragraphs": [ "Image input and general accuracy were not measured. Near-million-token needle retrieval does not prove general reasoning, decode speed or simultaneous long-session capacity. Runtime caches, Engram storage and per-node memory headroom need their own checks. Remaining slower states are acknowledged by the author." ], "citations": [ "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/docs/KNOWN-LIMITS.md" ] } ], "facts": [], "reported_results": [ { "label": "vllm-kit-run01 prose", "metric": "decode_rate", "value": 27.0, "unit": "tokens/s", "statistic": "one run; mean of two prompt-level decode rates; rounded by source", "uncertainty": null, "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "vllm-kit-run01", "engine": "vLLM + exllamav3 overlay", "engine_version": null, "runtime_execution_revision": null, "image_digest": "sha256:2f0cf3adc0f989c1d446be274df864eb799630175f604c3b22b71b7205971dce", "kit_revision": "6f7d1590ad49a2b8995188e45d7b9db31e677452", "draft_tokens": 3, "configured_max_requests": 1, "context_limit_tokens": 262144, "prefill_chunk_tokens": 1536, "chat_template": "kit template; actual CHAT_TEMPLATE override not read by the source", "workload": "prose", "per_prompt_decode_rates": [ 24.7, 29.3 ], "per_prompt_TTFT_seconds": [ 0.32, 0.28 ], "prompt_count": 2 }, "evidence_state": "source_reported", "site_reproduced": false, "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/01_vllm-miaai-kit_suite-results_32of32.json#L6", "source_document_sha256": "0cb7c61b5b321ba0b9eec76460e4d9d3d0e759bd09cfb03fc747ef66e7c2ea63", "source_document_id": "ds1017-runs-2026-10-17-01_vllm-miaai-kit_suite-results_32of32.json", "source_line": 6 }, { "label": "vllm-kit-run01 code", "metric": "decode_rate", "value": 39.9, "unit": "tokens/s", "statistic": "one run; mean of two prompt-level decode rates; rounded by source", "uncertainty": null, "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "vllm-kit-run01", "engine": "vLLM + exllamav3 overlay", "engine_version": null, "runtime_execution_revision": null, "image_digest": "sha256:2f0cf3adc0f989c1d446be274df864eb799630175f604c3b22b71b7205971dce", "kit_revision": "6f7d1590ad49a2b8995188e45d7b9db31e677452", "draft_tokens": 3, "configured_max_requests": 1, "context_limit_tokens": 262144, "prefill_chunk_tokens": 1536, "chat_template": "kit template; actual CHAT_TEMPLATE override not read by the source", "workload": "code", "per_prompt_decode_rates": [ 40.4, 39.4 ], "per_prompt_TTFT_seconds": [ 0.47, 0.32 ], "prompt_count": 2 }, "evidence_state": "source_reported", "site_reproduced": false, "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/01_vllm-miaai-kit_suite-results_32of32.json#L20", "source_document_sha256": "0cb7c61b5b321ba0b9eec76460e4d9d3d0e759bd09cfb03fc747ef66e7c2ea63", "source_document_id": "ds1017-runs-2026-10-17-01_vllm-miaai-kit_suite-results_32of32.json", "source_line": 20 }, { "label": "tensorfold-run11 prose", "metric": "decode_rate", "value": 56.9, "unit": "tokens/s", "statistic": "one run; mean of two prompt-level decode rates; rounded by source", "uncertainty": null, "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "tensorfold-run11", "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": "fd2365042f56fcf726cf4071132c8b779ad2eae0", "image_tag": "tensorfold-dsv41:fd2365042f56", "image_digest": null, "draft_tokens": 5, "configured_max_requests": 4, "context_limit_tokens": 1048576, "prefill_chunk_tokens": 2048, "chat_template": "patched kit template; source infers per-run use from file times", "workload": "prose", "per_prompt_decode_rates": [ 53.4, 60.4 ], "per_prompt_TTFT_seconds": [ 0.12, 0.12 ], "prompt_count": 2 }, "evidence_state": "source_reported", "site_reproduced": false, "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/11_tensorfold-fd2365042f56_suite-results_32of32.json#L6", "source_document_sha256": "d0ec8c99b0208ac4ddc272379d3cd71c796b5662d09481d6ba1e98fbbf0fa41c", "source_document_id": "ds1017-runs-2026-10-17-11_tensorfold-fd2365042f56_suite-results_32of32.json", "source_line": 6 }, { "label": "tensorfold-run11 code", "metric": "decode_rate", "value": 91.8, "unit": "tokens/s", "statistic": "one run; mean of two prompt-level decode rates; rounded by source", "uncertainty": null, "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-17", "concurrency": 1, "thinking": false, "temperature": 1.0, "top_p": 1.0, "max_output_tokens": 700, "output_tokens": null, "input_tokens": null, "prefix_cache": "present; cache-hit counts not logged in this suite", "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "(completion_tokens - 1) / (stream end - first chunk with content, reasoning or tool call)", "repetitions": 1, "checkpoint_is_latest_revision_verified": false, "id": "tensorfold-run11", "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": "fd2365042f56fcf726cf4071132c8b779ad2eae0", "image_tag": "tensorfold-dsv41:fd2365042f56", "image_digest": null, "draft_tokens": 5, "configured_max_requests": 4, "context_limit_tokens": 1048576, "prefill_chunk_tokens": 2048, "chat_template": "patched kit template; source infers per-run use from file times", "workload": "code", "per_prompt_decode_rates": [ 91.1, 92.5 ], "per_prompt_TTFT_seconds": [ 0.13, 0.16 ], "prompt_count": 2 }, "evidence_state": "source_reported", "site_reproduced": false, "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-17/11_tensorfold-fd2365042f56_suite-results_32of32.json#L20", "source_document_sha256": "d0ec8c99b0208ac4ddc272379d3cd71c796b5662d09481d6ba1e98fbbf0fa41c", "source_document_id": "ds1017-runs-2026-10-17-11_tensorfold-fd2365042f56_suite-results_32of32.json", "source_line": 20 }, { "label": "TensorFold final setup 32K prefill", "metric": "prefill_rate", "value": 2165, "unit": "tokens/s", "statistic": "median of five source rows; individually recoverable in per_run_values", "uncertainty": { "observed_min": 2065, "observed_max": 2201, "interval_kind": "observed range, not confidence interval" }, "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-18", "concurrency": 1, "thinking": false, "temperature": null, "top_p": null, "max_output_tokens": null, "output_tokens": null, "input_tokens": null, "prefix_cache": false, "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "prompt_tokens / TTFT; source final setup and host tuning", "repetitions": 5, "checkpoint_is_latest_revision_verified": false, "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": null, "runtime_revision_prefix": "92aa710b533d", "image_tag": "tensorfold-dsv41:92aa710b533d", "image_digest": null, "reported_prompt_bucket": "32K", "cached_tokens": 0, "per_run_values": [ 2201, 2180, 2065, 2165, 2100 ], "run_ids": [ "hw5", "hw6", "ht1", "ht2", "ht3" ], "draft_tokens": 5, "context_limit_tokens": 1048576, "warmup": "one 8K prompt; subsequent 8K requests before 32K/64K/128K; random-word prompts", "host_tuning_scope": "author applies no engine swap and memory-reclaim/compaction controls; not a universal host-tuning recommendation" }, "evidence_state": "source_reported", "site_reproduced": false, "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-18/INDEX.md#L16", "source_document_sha256": "e69d8c270e6407ddbaacb82da36cc48994b6f20fd4300d176753371b4e23fb8e", "source_document_id": "ds1017-runs-2026-10-18-INDEX.md", "source_line": 16 }, { "label": "TensorFold final setup 128K prefill", "metric": "prefill_rate", "value": 1961, "unit": "tokens/s", "statistic": "median of five source rows; individually recoverable in per_run_values", "uncertainty": { "observed_min": 1851, "observed_max": 1986, "interval_kind": "observed range, not confidence interval" }, "conditions": { "model": "DeepSeek-V4.1-Flash-1017", "model_artifact": "LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw", "model_revision": "afc7aa0aab98ee735c0f54e0a122019af1ac08f1", "Engram_artifact": "deepseek-ai/DeepSeek-V4.1-Flash-1017", "Engram_revision": "1bae0e1088993b9b82de08f59de5be5a34e42170", "weight_precision": "EXL3 2.9bpw", "system": "2 x NVIDIA DGX Spark GB10, TP2 over ConnectX-7", "node_count": 2, "driver": "580.178.04", "CUDA": "13.0", "experiment_date": "2026-10-18", "concurrency": 1, "thinking": false, "temperature": null, "top_p": null, "max_output_tokens": null, "output_tokens": null, "input_tokens": null, "prefix_cache": false, "kv_dtype": null, "kv_pool_tokens": null, "measurement_denominator": "prompt_tokens / TTFT; source final setup and host tuning", "repetitions": 5, "checkpoint_is_latest_revision_verified": false, "engine": "TensorFold", "engine_version": "0.6.6", "runtime_execution_revision": null, "runtime_revision_prefix": "92aa710b533d", "image_tag": "tensorfold-dsv41:92aa710b533d", "image_digest": null, "reported_prompt_bucket": "128K", "cached_tokens": 0, "per_run_values": [ 1986, 1981, 1914, 1961, 1851 ], "run_ids": [ "hw5", "hw6", "ht1", "ht2", "ht3" ], "draft_tokens": 5, "context_limit_tokens": 1048576, "warmup": "one 8K prompt; subsequent 8K requests before 32K/64K/128K; random-word prompts", "host_tuning_scope": "author applies no engine swap and memory-reclaim/compaction controls; not a universal host-tuning recommendation" }, "evidence_state": "source_reported", "site_reproduced": false, "evidence_url": "https://github.com/Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark/blob/9c6ab02340b770c6c3e1a9a4d5e70a56c5829d32/runs/2026-10-18/INDEX.md#L16", "source_document_sha256": "e69d8c270e6407ddbaacb82da36cc48994b6f20fd4300d176753371b4e23fb8e", "source_document_id": "ds1017-runs-2026-10-18-INDEX.md", "source_line": 16 } ], "source_document_sha256": "546a0c3bfbbb6e84226ec812cf38927f6d993e53b73633a540c426eea5f13d29" } }