Matches in SemOpenAlex for { <https://semopenalex.org/work/W2290640736> ?p ?o ?g. }
Showing items 1 to 65 of
65
with 100 items per page.
- W2290640736 abstract "Energy-efficient k-nearest-neighbor (kNN) computations are key building blocks for computer vision, classification, and machine-learning workloads [1–3]. Determining distances to high-dimensional vectors within a large vector database results in high compute cost. Adaptive precision improves energy efficiency by eliminating a majority of vectors without costly full-precision computation, with as-needed precision refinement to guarantee kNN accuracy of closely matched vectors. A special-purpose on-die kNN accelerator with 128-dimensions by 128 parallel reference vectors, targeted across mobile SoCs to multi-core microprocessors, and reconfigurable for either Manhattan or Euclidean distance, is fabricated in 14nm tri-gate CMOS [6]. Partial distance compute circuits, 2b window-based sort, MSB-to-LSB-based selective distance refinement, robust ultra-low voltage circuits, and state tracking control to selectively resume next-nearest candidates enable nominal energy efficiency of 3.37nJ/query vector or 9.7TOPS/W (measured for 21.5M vectors/s, 16 cycles/vector at 750mV, 25°C) with a dense layout occupying 0.333mm2 (Fig. 14.4.7) while achieving: i) scalable performance up to 26.4M vectors/s, 114mW measured at 850mV, ii) 2-cycle latency and 43pJ energy to find each subsequent nearest neighbor, iii) up to 5.2× higher throughput while maintaining full-precision kNN accuracy, iv) 16× search-space reduction for next-nearest neighbor, v) ultra-low voltage operation measured at 360mV, 1.1M vectors/s, 1.44mW, and vi) peak energy efficiency of 1.23nJ/vector at 390mV (near-threshold), 25°C." @default.
- W2290640736 created "2016-06-24" @default.
- W2290640736 creator A5006348328 @default.
- W2290640736 creator A5039276616 @default.
- W2290640736 creator A5052106795 @default.
- W2290640736 creator A5052890601 @default.
- W2290640736 creator A5070239387 @default.
- W2290640736 creator A5074107306 @default.
- W2290640736 creator A5075692568 @default.
- W2290640736 creator A5090166397 @default.
- W2290640736 date "2016-01-01" @default.
- W2290640736 modified "2023-10-14" @default.
- W2290640736 title "14.4 A 21.5M-query-vectors/s 3.37nJ/vector reconfigurable k-nearest-neighbor accelerator with adaptive precision in 14nm tri-gate CMOS" @default.
- W2290640736 cites W2003759031 @default.
- W2290640736 cites W2027066245 @default.
- W2290640736 cites W2119605622 @default.
- W2290640736 doi "https://doi.org/10.1109/isscc.2016.7418006" @default.
- W2290640736 hasPublicationYear "2016" @default.
- W2290640736 type Work @default.
- W2290640736 sameAs 2290640736 @default.
- W2290640736 citedByCount "9" @default.
- W2290640736 countsByYear W22906407362016 @default.
- W2290640736 countsByYear W22906407362017 @default.
- W2290640736 countsByYear W22906407362018 @default.
- W2290640736 countsByYear W22906407362019 @default.
- W2290640736 countsByYear W22906407362020 @default.
- W2290640736 countsByYear W22906407362022 @default.
- W2290640736 crossrefType "proceedings-article" @default.
- W2290640736 hasAuthorship W2290640736A5006348328 @default.
- W2290640736 hasAuthorship W2290640736A5039276616 @default.
- W2290640736 hasAuthorship W2290640736A5052106795 @default.
- W2290640736 hasAuthorship W2290640736A5052890601 @default.
- W2290640736 hasAuthorship W2290640736A5070239387 @default.
- W2290640736 hasAuthorship W2290640736A5074107306 @default.
- W2290640736 hasAuthorship W2290640736A5075692568 @default.
- W2290640736 hasAuthorship W2290640736A5090166397 @default.
- W2290640736 hasConcept C113238511 @default.
- W2290640736 hasConcept C121332964 @default.
- W2290640736 hasConcept C154945302 @default.
- W2290640736 hasConcept C41008148 @default.
- W2290640736 hasConcept C46362747 @default.
- W2290640736 hasConcept C49040817 @default.
- W2290640736 hasConceptScore W2290640736C113238511 @default.
- W2290640736 hasConceptScore W2290640736C121332964 @default.
- W2290640736 hasConceptScore W2290640736C154945302 @default.
- W2290640736 hasConceptScore W2290640736C41008148 @default.
- W2290640736 hasConceptScore W2290640736C46362747 @default.
- W2290640736 hasConceptScore W2290640736C49040817 @default.
- W2290640736 hasLocation W22906407361 @default.
- W2290640736 hasOpenAccess W2290640736 @default.
- W2290640736 hasPrimaryLocation W22906407361 @default.
- W2290640736 hasRelatedWork W1495317265 @default.
- W2290640736 hasRelatedWork W1999684908 @default.
- W2290640736 hasRelatedWork W206591153 @default.
- W2290640736 hasRelatedWork W2381236530 @default.
- W2290640736 hasRelatedWork W2381935838 @default.
- W2290640736 hasRelatedWork W2748952813 @default.
- W2290640736 hasRelatedWork W2899084033 @default.
- W2290640736 hasRelatedWork W2935759653 @default.
- W2290640736 hasRelatedWork W3012070291 @default.
- W2290640736 hasRelatedWork W3105167352 @default.
- W2290640736 isParatext "false" @default.
- W2290640736 isRetracted "false" @default.
- W2290640736 magId "2290640736" @default.
- W2290640736 workType "article" @default.