Steven Karl Reinhardt

This inventor holds 35 USPTO granted patents and 13 published patent applications and 4 EPO patents, primarily in Deep Learning Hardware. Top assignees: Advanced Micro Devices Corporation, Microsoft Technology Licensing, LLC, Intel Corporation. Active years: 2003-2024.

IDiyas Innovation Intelligence. (2026). Inventor Profile: Steven Karl Reinhardt. Retrieved from https://idiyas.com/inventor/steven-karl-reinhardt

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

USPTO Granted Patents = 35 (Patent Applications = 13)

% Patents Active = 80.0

EPO Patents = 4 (Patent Applications = 8)

Average Co-Inventor Count = 3.2

ph-index = 8

Forward Citations = 282(Granted Patents)

Location History:

Ann Arbor, MI (US) (2003 - 2009)
Ann Arbor, MA (US) (2009)
Bellevue, WA (US) (2021)
Vancouver, WA (US) (2007 - 2024)

Company Filing History:

13 out of 13,175

Advanced Micro Devices Corporation

patents

10 out of 56,279

Microsoft Technology Licensing, LLC

Hewlett-Packard Development Company, L.p.

patents

Years Active: 2003-2024

Loading Chart...

Areas of Expertise:

Deep Neural Networks

Memory Architecture

Accelerator Design

Tensor Processing

GPU Communication

Parallel Datapaths

Data Compression

Multithreading

Fault Detection

Instruction Set Architecture

Resource Virtualization

Cache Coherence

35 patents (USPTO):Sort by Forward CitationsExplore Patents

1. 12124391 - Semi-programmable and reconfigurable co-accelerator for a deep neural network with normalization or non-linearity (2024-10-22),

2. 12079137 - Banked memory architecture for multiple parallel datapath channels in an accelerator (2024-09-03),

3. 12073310 - Deep neural network accelerator with independent datapaths for simultaneous processing of different classes of operations (2024-08-27),

4. 11734214 - Semi-programmable and reconfigurable co-accelerator for a deep neural network with normalization or non-linearity (2023-08-22),

5. 11704251 - Banked memory architecture for multiple parallel datapath channels in an accelerator (2023-07-18),

6. 11347652 - Banked memory architecture for multiple parallel datapath channels in an accelerator (2022-05-31),

7. 10997116 - Tensor-based hardware accelerator including a scalar-processing unit (2021-05-04),

8. 10963309 - Network interface controller-based scheduling of processing tasks in a distributed computing system (2021-03-30),

9. 10936533 - GPU remote communication with triggered operations (2021-03-02),

10. 10372456 - Tensor processor instruction set architecture (2019-08-06),

11. 10360652 - Wavefront resource virtualization (2019-07-23),

12. 10338925 - Tensor register files (2019-07-02),

13. 10331445 - Multifunction vector processor circuits (2019-06-25),

14. 10320695 - Message aggregation, combining and compression for efficient data communications in GPU-based clusters (2019-06-11),

15. 10209990 - Conditional atomic operations in single instruction multiple data processors (2019-02-19),

Title: Innovator Spotlight: Steven Karl Reinhardt, Pioneering Technologist from Vancouver, WA

Introduction:

Steven Karl Reinhardt, a leading technologist hailing from Vancouver, WA, has emerged as a prominent figure in the realm of innovation and technology. With an extensive portfolio of 32 patents to his credit, Reinhardt's contributions have significantly influenced the fields of deep neural networks and memory architecture for accelerators.

Latest Patents:

Reinhardt's latest patents showcase his expertise in developing cutting-edge technologies:

1. Semi-programmable and reconfigurable co-accelerator for a deep neural network with normalization or non-linearity: This invention focuses on utilizing a configurable stacked architecture to accelerate operations or layers within a deep neural network. The architecture includes configurable micro-execution units that cater to various operations essential for DNN layers.

2. Banked memory architecture for multiple parallel datapath channels in an accelerator: This patent highlights the use of a banked memory structure to enhance the efficiency of accelerators. By segmenting dataflows and providing each channel with its memory bank, this innovation optimizes the performance of the accelerator.

Career Highlights:

Reinhardt's career is studded with significant milestones, having worked at esteemed companies such as Advanced Micro Devices Corporation and Intel Corporation. His tenure at these organizations played a pivotal role in shaping his expertise in technological advancements and patent-worthy innovations.

Collaborations:

Throughout his career, Reinhardt has collaborated with notable individuals in the industry. Some of his valued coworkers include Shubhendu Sekhar Mukherjee and Joel S Emer, with whom he has worked closely on various projects, further enriching his innovative endeavors.

Conclusion:

In conclusion, Steven Karl Reinhardt's exemplary work in the realm of technology and innovation underscores his visionary approach towards problem-solving and technological advancement. His patents and collaborations stand as a testament to his significant contributions to the field, cementing his position as a pioneering inventor and technologist in the industry.

Profile summary based on public USPTO records.