FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4376473245> ?p ?o ?g. }

• W4376473245 endingPage "37" @default.
• W4376473245 startingPage "19" @default.
• W4376473245 abstract "This paper presents an optimization method to build the smallest possible integer mapping unit that can replace a conventional multiply-and-accumulate unit in deep learning applications. The unit is built using a hardware-software co-design strategy that minimizes the set of represented real values and energy consumed. We target larger and more complex deep learning applications domains than those explored in previous related works, namely generative models for image and text content. Our key result is that using our proposed method, we can produce a set as small as 4 entries for an image enhancement application, and 16–32 entries for the GPT2 model, all with minimal loss of quality. Experimental results show that a hardware accelerator designed using our approach can reduce the processing time up to $$1.98times $$ / $$3.62times $$ and reduce computation energy consumed up to $$1.7times $$ / $$8.4times $$ compared to 8-bit integer/16-bit floating-point alternatives, respectively." @default.
• W4376473245 created "2023-05-14" @default.
• W4376473245 creator A5011045826 @default.
• W4376473245 creator A5056691858 @default.
• W4376473245 creator A5065637540 @default.
• W4376473245 creator A5080774688 @default.
• W4376473245 date "2023-01-01" @default.
• W4376473245 modified "2023-10-14" @default.
• W4376473245 title "Bedot: Bit Efficient Dot Product for Deep Generative Models" @default.
• W4376473245 cites W1791560514 @default.
• W4376473245 cites W1966812932 @default.
• W4376473245 cites W2047920195 @default.
• W4376473245 cites W2106291522 @default.
• W4376473245 cites W2133665775 @default.
• W4376473245 cites W2166029537 @default.
• W4376473245 cites W2289252105 @default.
• W4376473245 cites W2588762210 @default.
• W4376473245 cites W2766338242 @default.
• W4376473245 cites W2891158090 @default.
• W4376473245 cites W2954390672 @default.
• W4376473245 cites W2962793481 @default.
• W4376473245 cites W2963273111 @default.
• W4376473245 cites W3016829100 @default.
• W4376473245 cites W3034668968 @default.
• W4376473245 cites W3035083896 @default.
• W4376473245 cites W3103837983 @default.
• W4376473245 cites W3115518128 @default.
• W4376473245 cites W3137147200 @default.
• W4376473245 cites W3176714321 @default.
• W4376473245 cites W3177265267 @default.
• W4376473245 cites W3186740580 @default.
• W4376473245 cites W4206377087 @default.
• W4376473245 cites W4281657881 @default.
• W4376473245 cites W4285212630 @default.
• W4376473245 doi "https://doi.org/10.1007/978-3-031-32180-1_2" @default.
• W4376473245 hasPublicationYear "2023" @default.
• W4376473245 type Work @default.
• W4376473245 citedByCount "0" @default.
• W4376473245 crossrefType "book-chapter" @default.
• W4376473245 hasAuthorship W4376473245A5011045826 @default.
• W4376473245 hasAuthorship W4376473245A5056691858 @default.
• W4376473245 hasAuthorship W4376473245A5065637540 @default.
• W4376473245 hasAuthorship W4376473245A5080774688 @default.
• W4376473245 hasConcept C105795698 @default.
• W4376473245 hasConcept C108583219 @default.
• W4376473245 hasConcept C113775141 @default.
• W4376473245 hasConcept C11413529 @default.
• W4376473245 hasConcept C115961682 @default.
• W4376473245 hasConcept C154945302 @default.
• W4376473245 hasConcept C167966045 @default.
• W4376473245 hasConcept C177264268 @default.
• W4376473245 hasConcept C186370098 @default.
• W4376473245 hasConcept C199360897 @default.
• W4376473245 hasConcept C2524010 @default.
• W4376473245 hasConcept C26517878 @default.
• W4376473245 hasConcept C2777904410 @default.
• W4376473245 hasConcept C32900221 @default.
• W4376473245 hasConcept C33923547 @default.
• W4376473245 hasConcept C38652104 @default.
• W4376473245 hasConcept C39890363 @default.
• W4376473245 hasConcept C41008148 @default.
• W4376473245 hasConcept C45374587 @default.
• W4376473245 hasConcept C75695347 @default.
• W4376473245 hasConcept C84211073 @default.
• W4376473245 hasConcept C9390403 @default.
• W4376473245 hasConcept C97137487 @default.
• W4376473245 hasConceptScore W4376473245C105795698 @default.
• W4376473245 hasConceptScore W4376473245C108583219 @default.
• W4376473245 hasConceptScore W4376473245C113775141 @default.
• W4376473245 hasConceptScore W4376473245C11413529 @default.
• W4376473245 hasConceptScore W4376473245C115961682 @default.
• W4376473245 hasConceptScore W4376473245C154945302 @default.
• W4376473245 hasConceptScore W4376473245C167966045 @default.
• W4376473245 hasConceptScore W4376473245C177264268 @default.
• W4376473245 hasConceptScore W4376473245C186370098 @default.
• W4376473245 hasConceptScore W4376473245C199360897 @default.
• W4376473245 hasConceptScore W4376473245C2524010 @default.
• W4376473245 hasConceptScore W4376473245C26517878 @default.
• W4376473245 hasConceptScore W4376473245C2777904410 @default.
• W4376473245 hasConceptScore W4376473245C32900221 @default.
• W4376473245 hasConceptScore W4376473245C33923547 @default.
• W4376473245 hasConceptScore W4376473245C38652104 @default.
• W4376473245 hasConceptScore W4376473245C39890363 @default.
• W4376473245 hasConceptScore W4376473245C41008148 @default.
• W4376473245 hasConceptScore W4376473245C45374587 @default.
• W4376473245 hasConceptScore W4376473245C75695347 @default.
• W4376473245 hasConceptScore W4376473245C84211073 @default.
• W4376473245 hasConceptScore W4376473245C9390403 @default.
• W4376473245 hasConceptScore W4376473245C97137487 @default.
• W4376473245 hasLocation W43764732451 @default.
• W4376473245 hasOpenAccess W4376473245 @default.
• W4376473245 hasPrimaryLocation W43764732451 @default.
• W4376473245 hasRelatedWork W1974618110 @default.
• W4376473245 hasRelatedWork W2320349011 @default.
• W4376473245 hasRelatedWork W2379647085 @default.
• W4376473245 hasRelatedWork W2799397159 @default.
• W4376473245 hasRelatedWork W2947175736 @default.
• W4376473245 hasRelatedWork W3082497595 @default.

• next