Inventor profile of:

Daniel Lo

City:

Bothell, Washington

Country:

United States

Published Applications:

35

Last publication date:

2026-01-15

Top Assignees for applications by Daniel Lo

The entities that hold a legal rights for patent applications filed by inventor Lo Daniel:

Recent patent applications by Lo Daniel

Daniel Lo from Bothell, US has applied for patents for these inventions. The list has both pending applications and granted patents:

#1 | 2026-01-15
US20260017520A1
Physics

COMPRESSION AND STORAGE OF NEURAL NETWORK ACTIVATIONS FOR BACKPROPAGATION

#2 | 2025-02-20
US20250061320A1
Physics

ADJUSTING ACTIVATION COMPRESSION FOR NEURAL NETWORK TRAINING

#3 | 2024-08-15
US20240273287A1
Physics

HIGH-PERFORMANCE MICROCODED TEXT PARSER

#4 | 2024-05-09
US20240152758A1
Physics

Neural network activation compression with non-uniform mantissas

#5 | 2023-12-14
US20230403028A1
Electricity

PARALLELIZED DECODING OF VARIABLE-LENGTH PREFIX CODES

#6 | 2023-08-24
US20230267319A1
Physics

TRAINING NEURAL NETWORK ACCELERATORS USING MIXED PRECISION DATA FORMATS

#7 | 2023-06-22
US20230196085A1
Physics

RESIDUAL QUANTIZATION FOR NEURAL NETWORKS

#8 | 2023-05-04
US20230140185A1
Physics

Neural network activation compression with non-uniform mantissas

#9 | 2023-04-13
US20230110219A1
Physics

Neural network training with decreased memory consumption and processor utilization

#10 | 2022-09-15
US20220294447A1
Electricity

High-performance table-based state machine

#11 | 2022-03-17
US20220085815A1
Electricity

High-performance table-based state machine

#12 | 2022-03-17
US20220083732A1
Physics

High-performance microcoded text parser

#13 | 2021-12-28
US17084169
Electricity

Parallelized decoding of variable-length prefix codes

#14 | 2021-02-25
US20210056423A1
Physics

Neural network training with decreased memory consumption and processor utilization

#15 | 2020-09-24
US20200302330A1
Physics

Outlier quantization for training and inference

#16 | 2020-09-24
US20200302283A1
Physics

MIXED PRECISION TRAINING OF AN ARTIFICIAL NEURAL NETWORK

#17 | 2020-09-03
US20200279153A1
Physics

Deriving a concordant software neural network layer from a quantized firmware neural network layer

#18 | 2020-08-27
US20200272882A1
Physics

Neural network layer processing with normalization and transformation of data

#19 | 2020-08-27
US20200272881A1
Physics

Neural network layer processing with scaled quantization

#20 | 2020-08-20
US20200264876A1
Physics

Adjusting activation compression for neural network training

#21 | 2020-07-30
US20200242474A1
Physics

Neural network activation compression with non-uniform mantissas

#22 | 2020-07-09
US20200218982A1
Physics

DITHERED QUANTIZATION OF PARAMETERS DURING TRAINING WITH A MACHINE LEARNING TOOL

#23 | 2020-07-02
US20200210840A1
Physics

ADJUSTING PRECISION AND TOPOLOGY PARAMETERS FOR NEURAL NETWORK TRAINING BASED ON A PERFORMANCE METRIC

#24 | 2020-07-02
US20200210839A1
Physics

Neural network activation compression with outlier block floating-point

#25 | 2020-07-02
US20200210838A1
Physics

NEURAL NETWORK ACTIVATION COMPRESSION WITH NARROW BLOCK FLOATING-POINT

#26 | 2020-06-25
US20200202213A1
Physics

SCALED LEARNING FOR TRAINING DNN

#27 | 2020-06-18
US20200193274A1
Physics

Training neural network accelerators using mixed precision data formats

#28 | 2020-06-18
US20200193273A1
Physics

Residual quantization for neural networks

#29 | 2019-11-14
US20190347553A1
Physics

Training neural networks using mixed precision computations

#30 | 2019-11-14
US20190347072A1
Physics

Block floating point computations using shared exponents

#31 | 2019-11-07
US20190340499A1
Physics

QUANTIZATION FOR DNN ACCELERATORS

#32 | 2019-11-07
US20190340492A1
Physics

Flow for quantized neural networks

#33 | 2019-11-07
US20190339937A1
Physics

Block floating point computations using reduced bit-width vectors

#34 | 2019-09-19
US20190286973A1
Physics

HARDWARE ACCELERATED NEURAL NETWORK SUBGRAPHS

#35 | 2019-01-01
US15680649
Mechanical engineering

Hardware node having a matrix vector unit with block-floating point processing

InventorID:

2767493 ⎘