Paper Citation Record · LEDGER

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators

As of 21 August 2026, this Paper Citation Record lists 15 of 15 outbound references and 3 inbound Pith citation observations for arXiv:2506.09554.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2506.09554 v2

Coverage vector

measured 15 of 15 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T04:49:05.158865Z

measured 18 of 18 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-21T06:32:19.484+00:00

measured 3 of 3 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-06T11:07:58.959186Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: arxiv_reference, observed 2026-07-04T07:59:39.580222Z

Reference resolution

15 of 15 outbound references displayed

verified exact1
verified fuzzy8
unresolved4
parse uncertain0
malformed identifier2
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 8af69b5d-aa14-4aa8-a383-048688a79ac6 · outbound

This paper cites Evaluating Large Language Models Trained on Code.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Evaluating Large Language Models Trained on Code

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-07T04:49:03.534373Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9f5b5f4c-0136-46ec-8aab-02981a2b5328 · outbound

This paper cites Plug in the safety chip: Enforcing constraints for llm-driven robot agents,.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Plug in the safety chip: Enforcing constraints for llm-driven robot agents,

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:06.996503Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation 620bec5f-18e1-4327-a0c7-fcc3e9d3281c · outbound

This paper cites A Survey on Efficient Inference for Large Language Models.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators A Survey on Efficient Inference for Large Language Models

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-07T04:49:03.789514Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 89ecc77c-05b5-476d-8f20-896aca0967b4 · outbound

This paper cites LLM-Inference-Bench: Inference Benchmarking of Large Language Models on AI Accelerators.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators LLM-Inference-Bench: Inference Benchmarking of Large Language Models on AI Accelerators

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-07T04:49:03.916941Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b6633e4d-7b70-499f-83b9-4d4bc9029cc4 · outbound

This paper cites Characterizing the performance of accelerated jetson edge devices for training dnns,.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Characterizing the performance of accelerated jetson edge devices for training dnns,

Reference 5

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:06.859489Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation 550cf011-0485-43ba-8ed4-d1302fef8095 · outbound

This paper cites Large Language Models on Small Resource-Constrained Systems: Performance Characterization, Analysis and Trade-offs.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Large Language Models on Small Resource-Constrained Systems: Performance Characterization, Analysis and Trade-offs

Reference 6

Resolution: verified exact; local_arxiv, observed 2026-08-07T04:49:05.357455Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation 5952f333-940e-4bb0-8077-efa06d8ea389 · outbound

This paper cites A preliminary performance analysis of llm inference on edge accelerators,.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators A preliminary performance analysis of llm inference on edge accelerators,

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:06.675928Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation 90bfe4cf-22eb-4e00-989f-2a49948496f8 · outbound

This paper cites Wikitext-2,.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Wikitext-2,

Reference 8

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:06.518047Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation 32235912-1aff-4d2f-b2b7-0818a9672b08 · outbound

This paper cites Longbench,.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Longbench,

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:06.334059Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation 37ed37ed-a5e6-4ed3-a6ea-ef8ce45a6286 · outbound

This paper cites Gpt3.int8(): 8-bit matrix multiplication for transformers at scale,.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Gpt3.int8(): 8-bit matrix multiplication for transformers at scale,

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:06.160849Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation 8315a6d7-71c0-4c55-ab46-ddabf1f332ec · outbound

This paper cites Splitwise: Efficient generative LLM inference using phase splitting.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Splitwise: Efficient generative LLM inference using phase splitting

Reference 11

Resolution: malformed identifier; no resolver link, observed 2026-08-07T04:49:04.761986Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0f1a1851-5edd-4ea6-b0f6-8c6664812145 · outbound

This paper cites When compared with INT4, INT8’s power savings range between 20% and 43% (median 32%).

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators When compared with INT4, INT8’s power savings range between 20% and 43% (median 32%)

Reference 12

Resolution: malformed identifier; raw_fallback, observed 2026-08-07T04:49:06.022150Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation cc80d7e8-585a-4c02-b903-569a8cb2d6b3 · outbound

This paper cites Against INT4, INT8 consistently yields over 27% power savings.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Against INT4, INT8 consistently yields over 27% power savings

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:05.886948Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation cb242ee6-3e20-4f56-bead-44e9b01bc9a0 · outbound

This paper cites •Energy Consumption: Similarly, INT8 achieves lower energy usage, with a median reduction of 24% compared to FP16 and 55% com- pared to INT4.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators •Energy Consumption: Similarly, INT8 achieves lower energy usage, with a median reduction of 24% compared to FP16 and 55% com- pared to INT4

Reference 14

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T04:49:05.737086Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation dfb5deec-c2a4-4ee6-b37a-dce61653e0fb · outbound

This paper cites an unresolved cited work.

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators Unresolved cited work

Reference 15

Resolution: unresolved; raw_fallback, observed 2026-08-07T04:49:05.557627Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Pith citing papers

Observation 09e339dd-26dd-4613-8d3b-302929859ad7 · inbound

On the Sustainability of AI Inferences in the Edge cites this paper.

On the Sustainability of AI Inferences in the Edge Understanding the Performance and Power of LLM Inferencing on Edge Accelerators

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-06T11:07:58.959186Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 056f7926-2373-437c-93e6-ce0d485aa266 · inbound

Does Mixture-of-Experts Actually Help Inference on Consumer and Edge Hardware? An Empirical Study cites this paper.

Does Mixture-of-Experts Actually Help Inference on Consumer and Edge Hardware? An Empirical Study Understanding the Performance and Power of LLM Inferencing on Edge Accelerators

Reference 2

Resolution: verified exact; arxiv_id, observed 2026-07-04T07:59:39.582044Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.

Observation bfb30276-44c1-4ed4-866f-de8350740453 · inbound

Does Mixture-of-Experts Actually Help Inference on Consumer and Edge Hardware? An Empirical Study cites this paper.

Does Mixture-of-Experts Actually Help Inference on Consumer and Edge Hardware? An Empirical Study Understanding the Performance and Power of LLM Inferencing on Edge Accelerators

Reference 2

Resolution: unresolved; no resolver link, observed 2026-07-12T13:05:17.273287Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.