Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313000323> ?p ?o ?g. }
Showing items 1 to 92 of
92
with 100 items per page.
- W4313000323 endingPage "1269" @default.
- W4313000323 startingPage "1256" @default.
- W4313000323 abstract "Advanced edge sensing/computing devices, such as AR/VR devices, have a uniquely challenging adaptive baseline workload and camera sensor structure. These devices must process images in real-time from multiple sensors, placing a large burden on a typical centralized mobile SoC processor. Augmenting the sensors with a package-integrated near-sensor processor can improve the device’s processing performance as well as reduce energy consumption. This near-sensor processor must adapt to the dynamic workloads, fit within a limited silicon footprint and energy envelope, and satisfy the real-time requirement. In this work, we present ANSA, a near-sensor processor architecture supporting flexible processing schemes and dataflows to maintain high efficiency for dynamic CNN workloads. ANSA is scalable to sub-mm2 sizes to match the footprint of advanced image sensors. ANSA supports module-level power gating to adapt the compute capacity to dynamic workloads. Finally, ANSA leverages recent advancements in high-density non-volatile memory and 3D packaging to support weight storage within the area constraints of an image sensor. Overall, ANSA achieves inference energy consumption up to <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$30times $ </tex-math></inline-formula> lower than a standard SIMD baseline. Additionally, our design’s scalability allows it to achieve up to <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$2.76times $ </tex-math></inline-formula> lower average inference energy at <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$4.5times $ </tex-math></inline-formula> lower silicon area compared to competing edge accelerator designs." @default.
- W4313000323 created "2023-01-05" @default.
- W4313000323 creator A5016814203 @default.
- W4313000323 creator A5038110244 @default.
- W4313000323 creator A5053194542 @default.
- W4313000323 creator A5072004972 @default.
- W4313000323 creator A5086838214 @default.
- W4313000323 date "2023-03-01" @default.
- W4313000323 modified "2023-10-16" @default.
- W4313000323 title "ANSA: Adaptive Near-Sensor Architecture for Dynamic DNN Processing in Compact Form Factors" @default.
- W4313000323 cites W2029990606 @default.
- W4313000323 cites W2073818373 @default.
- W4313000323 cites W2109351996 @default.
- W4313000323 cites W2442974303 @default.
- W4313000323 cites W2606722458 @default.
- W4313000323 cites W2606794968 @default.
- W4313000323 cites W2612076670 @default.
- W4313000323 cites W2787701666 @default.
- W4313000323 cites W2790925711 @default.
- W4313000323 cites W2793847918 @default.
- W4313000323 cites W2883929540 @default.
- W4313000323 cites W2895738878 @default.
- W4313000323 cites W2899481200 @default.
- W4313000323 cites W2921126577 @default.
- W4313000323 cites W2945146780 @default.
- W4313000323 cites W2963125977 @default.
- W4313000323 cites W2964988320 @default.
- W4313000323 cites W2979838629 @default.
- W4313000323 cites W2980104813 @default.
- W4313000323 cites W2982083293 @default.
- W4313000323 cites W3006200321 @default.
- W4313000323 cites W3025341614 @default.
- W4313000323 cites W3048746542 @default.
- W4313000323 cites W3091166046 @default.
- W4313000323 cites W3107849462 @default.
- W4313000323 cites W3133687618 @default.
- W4313000323 cites W4240168186 @default.
- W4313000323 cites W4249932213 @default.
- W4313000323 cites W4253012315 @default.
- W4313000323 doi "https://doi.org/10.1109/tcsi.2022.3228725" @default.
- W4313000323 hasPublicationYear "2023" @default.
- W4313000323 type Work @default.
- W4313000323 citedByCount "0" @default.
- W4313000323 crossrefType "journal-article" @default.
- W4313000323 hasAuthorship W4313000323A5016814203 @default.
- W4313000323 hasAuthorship W4313000323A5038110244 @default.
- W4313000323 hasAuthorship W4313000323A5053194542 @default.
- W4313000323 hasAuthorship W4313000323A5072004972 @default.
- W4313000323 hasAuthorship W4313000323A5086838214 @default.
- W4313000323 hasConcept C113775141 @default.
- W4313000323 hasConcept C118524514 @default.
- W4313000323 hasConcept C119599485 @default.
- W4313000323 hasConcept C127413603 @default.
- W4313000323 hasConcept C149635348 @default.
- W4313000323 hasConcept C2780165032 @default.
- W4313000323 hasConcept C41008148 @default.
- W4313000323 hasConcept C48044578 @default.
- W4313000323 hasConcept C7366592 @default.
- W4313000323 hasConcept C77088390 @default.
- W4313000323 hasConcept C9390403 @default.
- W4313000323 hasConceptScore W4313000323C113775141 @default.
- W4313000323 hasConceptScore W4313000323C118524514 @default.
- W4313000323 hasConceptScore W4313000323C119599485 @default.
- W4313000323 hasConceptScore W4313000323C127413603 @default.
- W4313000323 hasConceptScore W4313000323C149635348 @default.
- W4313000323 hasConceptScore W4313000323C2780165032 @default.
- W4313000323 hasConceptScore W4313000323C41008148 @default.
- W4313000323 hasConceptScore W4313000323C48044578 @default.
- W4313000323 hasConceptScore W4313000323C7366592 @default.
- W4313000323 hasConceptScore W4313000323C77088390 @default.
- W4313000323 hasConceptScore W4313000323C9390403 @default.
- W4313000323 hasFunder F4320319290 @default.
- W4313000323 hasIssue "3" @default.
- W4313000323 hasLocation W43130003231 @default.
- W4313000323 hasOpenAccess W4313000323 @default.
- W4313000323 hasPrimaryLocation W43130003231 @default.
- W4313000323 hasRelatedWork W1581492067 @default.
- W4313000323 hasRelatedWork W1975010174 @default.
- W4313000323 hasRelatedWork W2019641810 @default.
- W4313000323 hasRelatedWork W2152033542 @default.
- W4313000323 hasRelatedWork W2383828164 @default.
- W4313000323 hasRelatedWork W2392817843 @default.
- W4313000323 hasRelatedWork W2563665962 @default.
- W4313000323 hasRelatedWork W2784141701 @default.
- W4313000323 hasRelatedWork W3120511008 @default.
- W4313000323 hasRelatedWork W4246246179 @default.
- W4313000323 hasVolume "70" @default.
- W4313000323 isParatext "false" @default.
- W4313000323 isRetracted "false" @default.
- W4313000323 workType "article" @default.