MétaCan
Menu
Back to cohort

Systematic Framework Leveraging Early Assessment of DNN Reliability for Efficient and Reliable FPGA-Based Inference Deployment

2025· article· W4417169891 on OpenAlexaff
Vu Trinh, Otmane Aı̈t Mohamed, Fakhreddine Ghaffari

Bibliographic record

Venuenot available
Typearticle
Language
FieldComputer Science
TopicAdvanced Neural Network Applications
Canadian institutionsConcordia University
Fundersnot available
KeywordsInferenceSoftware deploymentPython (programming language)Field-programmable gate arrayArtificial neural networkSoftwareReliability (semiconductor)Scheduling (production processes)Fault tolerance

Abstract

fetched live from OpenAlex

Deep Neural Networks (DNNs) deployed on edge AI hardware accelerators are increasingly susceptible to transient faults caused by environmental disturbances, often resulting in soft errors that degrade system reliability. In this work, we present a comprehensive and extensible framework that enables early evaluation of DNN reliability purely at the Python software level, supporting arbitrary user-defined models and datasets. The framework systematically incorporates the effects of aforementioned optimization techniques and introduces integrated error mitigation strategies to assess their effectiveness pre-hardware deployment. To bridge software-level evaluation with hardware implementation, we extend our framework to a hardware/software codesign platform featuring a configurable systolic array on FPGA. This array, enhanced with built-in fault mitigation mechanisms, accelerates convolution and linear layers, while an ARM CPU core completes the remaining inference tasks. The proposed system achieves resilient DNN inference with a 2.2% increase in FPGA LUT utilization (for LeNet-5 on MNIST). By enabling early-stage analysis and optimization, our approach facilitates robust and resource-efficient DNN deployment for edge AI applications.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.001
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow)
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Methods · Consensus signal: none
Teacher disagreement score0.589
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0000.002
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0010.001
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.021
GPT teacher head0.326
Teacher spread0.305 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

Study designSimulation or modeling
Domainnot available
GenreMethods

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same topicAdvanced Neural Network ApplicationsFrench-language works237,207