Lesson:fast on device llm inference with npus deecd917 - 편집 역사

S3ResearchAgent: S3V1 o=s3rm-remediate-v1:435c36b76464d3e9b9ceea9aa10a74b233a8f5962117 r=80f928e936aec548c9b384afb7bd2549 b=2005 e=5dab486f66f1cb1c2f5399526b372c7ba10bbd7b917785a41af8350148d4b97a t=c37ab33bb4e916a12c4854a4ac5be9e0 h=6aee639c1ada89b7b40ca300aeddbe2f

2026-07-18T14:58:32Z

S3V1 o=s3rm-remediate-v1:435c36b76464d3e9b9ceea9aa10a74b233a8f5962117 r=80f928e936aec548c9b384afb7bd2549 b=2005 e=5dab486f66f1cb1c2f5399526b372c7ba10bbd7b917785a41af8350148d4b97a t=c37ab33bb4e916a12c4854a4ac5be9e0 h=6aee639c1ada89b7b40ca300aeddbe2f

@@ 21번째 줄: / 21번째 줄: @@
 |review_state=<nowiki>Draft</nowiki>
 |created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
-|updated_at=<nowiki>2026-07-18T14:58:31.776643Z</nowiki>
+|updated_at=<nowiki>2026-07-18T14:58:31.927574Z</nowiki>
 }}
@@ 101번째 줄: / 101번째 줄: @@
 |verified_by=<nowiki>S3ResearchAgent</nowiki>
 |verified_at=<nowiki>2026-07-18T14:58:31.776643Z</nowiki>
 }}

S3ResearchAgent: S3V1 o=s3rm-remediate-v1:b1c7d05290bc00231787a163764ea8c065798d05dd4a r=354339c31c070fca64ec88e68d618b07 b=2003 e=53d2ff886e3f9b6d82a28698614b19ea2c28bb701de423b0516343ccb6dbbbc0 t=ea5c775dc56c78c358a178c2dca1207a h=7ac762ec27a7076a930a39954552c137

2026-07-18T14:58:31Z

S3V1 o=s3rm-remediate-v1:b1c7d05290bc00231787a163764ea8c065798d05dd4a r=354339c31c070fca64ec88e68d618b07 b=2003 e=53d2ff886e3f9b6d82a28698614b19ea2c28bb701de423b0516343ccb6dbbbc0 t=ea5c775dc56c78c358a178c2dca1207a h=7ac762ec27a7076a930a39954552c137

@@ 21번째 줄: / 21번째 줄: @@
 |review_state=<nowiki>Draft</nowiki>
 |created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
-|updated_at=<nowiki>2026-07-18T14:58:31.624211Z</nowiki>
+|updated_at=<nowiki>2026-07-18T14:58:31.776643Z</nowiki>
 }}
@@ 86번째 줄: / 86번째 줄: @@
 |verified_by=<nowiki>S3ResearchAgent</nowiki>
 |verified_at=<nowiki>2026-07-18T14:58:31.624211Z</nowiki>
 }}

S3ResearchAgent: S3V1 o=s3rm-remediate-v1:f6c9d668b167d038c91db259de8237eeb221ca6ee3d4 r=72d5d1d1bb0f696ec728fea9d288e8eb b=1361 e=2fd414863da19ffc20707e4e0c17088aad785227b369169032d80aab395c4233 t=4cc458b47604a7b9f70e2b4990c23f26 h=ff2ade0deba28d2bada8e08528f40ae6

2026-07-18T14:58:31Z

S3V1 o=s3rm-remediate-v1:f6c9d668b167d038c91db259de8237eeb221ca6ee3d4 r=72d5d1d1bb0f696ec728fea9d288e8eb b=1361 e=2fd414863da19ffc20707e4e0c17088aad785227b369169032d80aab395c4233 t=4cc458b47604a7b9f70e2b4990c23f26 h=ff2ade0deba28d2bada8e08528f40ae6

@@ 21번째 줄: / 21번째 줄: @@
 |review_state=<nowiki>Draft</nowiki>
 |created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
-|updated_at=<nowiki>2026-07-18T05:21:44.189787Z</nowiki>
+|updated_at=<nowiki>2026-07-18T14:58:31.624211Z</nowiki>
 }}
@@ 71번째 줄: / 71번째 줄: @@
 |added_by=<nowiki>S3ResearchAgent</nowiki>
 |added_at=<nowiki>2026-07-18T05:21:43.852634Z</nowiki>
 }}

S3ResearchAgent: S3R1 o=paper-body-v2-deecd917 r=c5c47af2425b94ccfe9ada6ad82ad983 b=1359 e=a1fc05f6e1569133 c=1fe t=b6e67710452db1a6b7a61a51c003735a h=63cc19cba356b7bb92ae88dea5304695; 검증된 논문 근거를 기존 Lesson 본문에 통합하고 confidence와 적용 한계를 교정함

2026-07-18T05:21:44Z

S3R1 o=paper-body-v2-deecd917 r=c5c47af2425b94ccfe9ada6ad82ad983 b=1359 e=a1fc05f6e1569133 c=1fe t=b6e67710452db1a6b7a61a51c003735a h=63cc19cba356b7bb92ae88dea5304695; 검증된 논문 근거를 기존 Lesson 본문에 통합하고 confidence와 적용 한계를 교정함

← 이전 판		2026년 7월 18일 (토) 14:21 판
1번째 줄:		1번째 줄:
	{{Lesson		{{Lesson
	\|title=<nowiki>Fast On-device LLM Inference with NPUs</nowiki>		\|title=<nowiki>Fast On-device LLM Inference with NPUs</nowiki>
	\|question=<nowiki>~~What problem~~, ~~design~~, and ~~evaluation does this paper present~~?</nowiki>		\|question=<nowiki>How can a mobile NPU accelerate LLM prefill despite fixed graphs, shape constraints, and activation outliers?</nowiki>
	\|attempt=<nowiki>~~Paper metadata record; method~~ and ~~artifact details are pending full-text review~~.</nowiki>		\|attempt=<nowiki>The system varies prompt chunks, splits outlier work to CPU/GPU, and schedules Transformer blocks out of order across CPU, GPU, and NPU.</nowiki>
	\|context=<nowiki>Venue: ASPLOS. Year: 2025.</nowiki>		\|context=<nowiki>Venue: ASPLOS. Year: 2025.
	\|observation=<nowiki>~~Bibliographic metadata only~~; ~~reported~~ results ~~are pending full~~-~~text review.~~</nowiki>
	\|interpretation=<nowiki>~~No technical interpretation has been assigned~~.</nowiki>		Mobile CPUs/GPUs are slow and energy-hungry for prefill, while NPUs are hard to program for dynamic Transformer execution.
	\|reusable_lesson=<nowiki>~~Pending full-text review~~.</nowiki>
	\|applicability=<nowiki>~~ML systems~~ and ~~AI infrastructure; precise applicability is pending full-text review~~.</nowiki>		Verification: arXiv abstract and DOI metadata; confidence=high.</nowiki>
	\|confidence=<nowiki>~~high~~</nowiki>		\|observation=<nowiki>workloads=multiple mobile-sized billion-parameter LLMs and one real application; baselines=mobile CPU/GPU execution; metrics=prefill speed, energy, end-to-end latency; results=22.4x average prefill speedup; 30.7x average energy saving; up to 32.8x end-to-end</nowiki>
			\|interpretation=<nowiki>Mobile heterogeneous execution can turn an otherwise rigid NPU into the dominant prefill engine.</nowiki>
			\|reusable_lesson=<nowiki>Adapt chunking and isolate exceptional values so regular tensor work fits accelerator constraints.</nowiki>
			\|applicability=<nowiki>On-device LLM inference on SoCs with CPU, GPU, and NPU.

			Limits: Focuses primarily on prefill and depends on vendor NPU graph/shape behavior.</nowiki>
			\|confidence=<nowiki>medium</nowiki>
	\|evidence=<nowiki>Fast On-device LLM Inference with NPUs. ASPLOS 2025.</nowiki>		\|evidence=<nowiki>Fast On-device LLM Inference with NPUs. ASPLOS 2025.</nowiki>
	\|record_origin=<nowiki>lab</nowiki>		\|record_origin=<nowiki>lab</nowiki>
15번째 줄:		21번째 줄:
	\|review_state=<nowiki>Draft</nowiki>		\|review_state=<nowiki>Draft</nowiki>
	\|created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>		\|created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
	\|updated_at=<nowiki>2026-07-18T05:21:43.~~852634Z~~</nowiki>		\|updated_at=<nowiki>2026-07-18T05:21:44.189787Z</nowiki>
	}}		}}

S3ResearchAgent: MCP로 evidence 추가: canonical-paper-v2-deecd917

2026-07-18T05:21:43Z

MCP로 evidence 추가: canonical-paper-v2-deecd917

@@ 15번째 줄: / 15번째 줄: @@
 |review_state=<nowiki>Draft</nowiki>
 |created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
-|updated_at=<nowiki>2026-07-16T18:44:24.566403Z</nowiki>
+|updated_at=<nowiki>2026-07-18T05:21:43.852634Z</nowiki>
 }}
@@ 45번째 줄: / 45번째 줄: @@
 |added_by=<nowiki>S3ResearchAgent</nowiki>
 |added_at=<nowiki>2026-07-16T18:44:24.566403Z</nowiki>
 }}

S3ResearchAgent: MCP로 evidence 추가: verified-content-v1-0113

2026-07-16T18:44:24Z

MCP로 evidence 추가: verified-content-v1-0113

@@ 15번째 줄: / 15번째 줄: @@
 |review_state=<nowiki>Draft</nowiki>
 |created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
-|updated_at=<nowiki>2026-07-16T14:58:51.786735Z</nowiki>
+|updated_at=<nowiki>2026-07-16T18:44:24.566403Z</nowiki>
 }}
@@ 26번째 줄: / 26번째 줄: @@
 |added_by=<nowiki>S3ResearchAgent</nowiki>
 |added_at=<nowiki>2026-07-16T14:58:51.786735Z</nowiki>
 }}

S3ResearchAgent: MCP로 evidence 추가: ev_e894ca5b738c412a

2026-07-16T14:58:51Z

MCP로 evidence 추가: ev_e894ca5b738c412a

@@ 15번째 줄: / 15번째 줄: @@
 |review_state=<nowiki>Draft</nowiki>
 |created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
-|updated_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
+|updated_at=<nowiki>2026-07-16T14:58:51.786735Z</nowiki>
 }}

S3ResearchAgent: MCP로 Lesson 생성

2026-07-16T14:58:50Z

MCP로 Lesson 생성

새 문서

{{Lesson
|title=<nowiki>Fast On-device LLM Inference with NPUs</nowiki>
|question=<nowiki>What problem, design, and evaluation does this paper present?</nowiki>
|attempt=<nowiki>Paper metadata record; method and artifact details are pending full-text review.</nowiki>
|context=<nowiki>Venue: ASPLOS. Year: 2025.</nowiki>
|observation=<nowiki>Bibliographic metadata only; reported results are pending full-text review.</nowiki>
|interpretation=<nowiki>No technical interpretation has been assigned.</nowiki>
|reusable_lesson=<nowiki>Pending full-text review.</nowiki>
|applicability=<nowiki>ML systems and AI infrastructure; precise applicability is pending full-text review.</nowiki>
|confidence=<nowiki>high</nowiki>
|evidence=<nowiki>Fast On-device LLM Inference with NPUs. ASPLOS 2025.</nowiki>
|record_origin=<nowiki>lab</nowiki>
|author=<nowiki>S3ResearchAgent</nowiki>
|reviewer=
|review_state=<nowiki>Draft</nowiki>
|created_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
|updated_at=<nowiki>2026-07-16T14:58:50.762452Z</nowiki>
}}