Inventor · Bothell, WA, US

Daniel Lo

28Patents
3h-index
31Co-inventors
63Inventor score

Filing activity: Mar 2, 1989 → Jan 17, 2024

Most-cited inventions

PatentTitleAreaCited byStatus
US5155837A Methods and apparatus for software retrofitting Electricity 194 Expired
US10167800B1 Hardware node having a matrix vector unit with block-floating point processing Physics 71 Active
US5707014A Water cleaning apparatus Performing Operations; Transporting 20 Expired
US10241970B2 Reduced memory nucleotide sequence comparison Physics 1 Active
US11526761B2 Neural network training with decreased memory consumption and processor utilization Physics 1 Active
US11586883B2 Residual quantization for neural networks Physics 1 Active
US11381241B2 High-performance table-based state machine Electricity 0 Active
US11863182B2 High-performance table-based state machine Electricity 0 Active
US11741362B2 Training neural networks using mixed precision computations Electricity 0 Active
US11562201B2 Neural network layer processing with normalization and transformation of data Physics 0 Active
US11562247B2 Neural network activation compression with non-uniform mantissas Electricity 0 Active
US11211945B1 Parallelized decoding of variable-length prefix codes Electricity 0 Active
US11556764B2 Deriving a concordant software neural network layer from a quantized firmware neural network layer Physics 0 Active
US11989508B2 High-performance microcoded text parser Electricity 0 Active
US12045724B2 Neural network activation compression with outlier block floating-point Electricity 0 Active
US11544521B2 Neural network layer processing with scaled quantization Physics 0 Active
US12277502B2 Neural network activation compression with non-uniform mantissas Electricity 0 Active
US10691413B2 Block floating point computations using reduced bit-width vectors Physics 0 Active
US10579334B2 Block floating point computations using shared exponents Physics 0 Active
US12067495B2 Neural network activation compression with non-uniform mantissas Electricity 0 Active
US11676003B2 Training neural network accelerators using mixed precision data formats Physics 0 Active
US11853897B2 Neural network training with decreased memory consumption and processor utilization Physics 0 Active
US11645493B2 Flow for quantized neural networks Physics 0 Active
US12426036B2 Mapping entities to accounts for de-anonymization of online activity Electricity 0 Active
US10566076B2 Customized integrated circuit for serial comparison of nucleotide sequences Physics 0 Active

Source: USPTO / EPO open patent data. Inventor disambiguation is heuristic; counts are objective bibliographic measures.