logo

For the Inventor, By the Inventor

San Jose, CA, United States of America

HsinYu Tsai

This inventor holds 17 USPTO granted patents and 5 published patent applications and 1 EPO patent, primarily in Analog Artificial Intelligence. Top assignees: International Business Machines Corporation, International Business Machines Cororation. Active years: 2019-2026.

IDiyas Innovation Intelligence. (2026). Inventor Profile: HsinYu Tsai. Retrieved from https://idiyas.com/inventor/hsinyu-tsai-g5t6tany

Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile

Calculating rank
USPTO Granted Patents = 17 

% Patents Active = 58.8

 

Average Co-Inventor Count = 3.9

ph-index = 2

Forward Citations = 10(Granted Patents)


Company Filing History:

goldMedal16 out of 166,425 
 
International Business Machines Corporation
 patents
silverMedal1 out of 1 
 
International Business Machines Cororation
 patent

Years Active: 2019-2026

Loading Chart...
Loading Chart...
Areas of Expertise:
Calibrating Peripheral Variability
Analog Resistive Processing Unit
Neural Network Accelerators
Compute-In-Memory Architecture
Row-By-Row Convolutional Mapping
Drift Regularization
Phase-Change Memory
Copper Patterning
Magnetoresistive Random-Access Memory
Biosensor Optimization
III-V CMOS Co-Integration
Self-Assembly Techniques
17 patents (USPTO):Explore Patents
1. 12626111 - Calibrating peripheral variability (2026-05-12),
2. 12585940 - Learning static bound management parameters for analog resistive processing unit system (2026-03-24),
3. 12518149 - Implicit vector concatenation within 2D mesh routing (2026-01-06),
4. 12456043 - Two-dimensional mesh for compute-in-memory accelerator architecture (2025-10-28),
5. 12229680 - Neural network accelerators resilient to conductance drift (2025-02-18),
6. 12050997 - Row-by-row convolutional neural network mapping for analog artificial intelligence network training (2024-07-30),
7. 11868893 - Efficient tile mapping for row-by-row convolutional neural network mapping for analog artificial intelligence network inference (2024-01-09),
8. 11562240 - Efficient tile mapping for row-by-row convolutional neural network mapping for analog artificial intelligence network inference (2023-01-24),
9. 11514326 - Drift regularization to counteract variation in drift coefficients for analog accelerators (2022-11-29),
10. 11461640 - Mitigation of conductance drift in neural network resistive processing units (2022-10-04),
11. 11347999 - Closed loop programming of phase-change memory (2022-05-31),
12. 10930565 - III-V CMOS co-integration (2021-02-23),
13. 10665325 - Reduction of surface nucleotide hybridization by optimizing a biosensor sensing surface area (2020-05-26),
14. 10600656 - Directed self-assembly for copper patterning (2020-03-24),
15. 10593870 - Sidewall image transfer on magnetic tunnel junction stack for magnetoresistive random-access memory patterning (2020-03-17),

Title: HsinYu Tsai: Innovator in Compute-in-Memory Accelerator Architecture

Introduction

HsinYu Tsai is a prominent inventor based in San Jose, California, known for his significant contributions to the field of compute-in-memory (CIM) technology. With a total of 14 patents to his name, Tsai has been at the forefront of innovations that enhance the performance of deep neural networks.

Latest Patents

Among his latest patents, Tsai has developed a two-dimensional mesh for compute-in-memory accelerator architecture. This architecture is designed for deep neural networks and includes a first analog fabric engine with multiple compute-in-memory analog tiles. Each tile is capable of storing a matrix of weight operands and performing in-memory computations to produce outputs from inputs. Additionally, he has patented a method for creating neural network accelerators that are resilient to conductance drift. This method involves applying an arbitrary amplification factor to the activation function outputs of neurons, compensating for variations in resistive processing unit weights.

Career Highlights

HsinYu Tsai has had a distinguished career, working with notable companies such as IBM. His work has focused on advancing technologies that improve computational efficiency and accuracy in neural networks.

Collaborations

Throughout his career, Tsai has collaborated with esteemed colleagues, including Pritish Narayanan and Geoffrey W. Burr. These collaborations have contributed to the development of innovative solutions in the field of computing.

Conclusion

HsinYu Tsai's contributions to compute-in-memory technology and neural networks have positioned him as a leading inventor in his field. His innovative patents and collaborations continue to influence advancements in technology.

Profile summary based on public USPTO records.
Data Sources: USPTO Patent Grant XML, Patent Center, EPO & CIPO • Normalized by IDiyas Innovation Graph. Methodology & provenance architecturePlease report any incorrect information to [email protected]
Loading…