PATENT PORTFOLIO · UPDATED AUGUST 2026
Inventions in memory architecture and AI systems.
Patents spanning high-bandwidth memory, processing-in-memory, GPGPUs, and AI systems.
13U.S. patent grants
3international grants
4published pending U.S. applications
Filter by theme
Organize by
Showing 20 of 20 patents and applications
01
U.S. patent grants
132026July 21, 2026
Arithmetic logic unit in a base die of a processing-in-memory component with cross-ALU data communication capability
US 12,688,123↗
2026June 23, 2026
Data encryption suitable for use in systems with processing-in-memory
US 12,664,319↗
2026June 16, 2026
3D layout and organization for enhancement of modern memory systems
US 12,660,703↗
2026March 10, 2026
Memory controller and near-memory support for sparse accesses
US 12,572,280↗
2026January 6, 2026
Method and apparatus for recovering regular access performance in fine-grained DRAM
US 12,518,819↗
2025September 16, 2025
Transforming all-bank processing-in-memory operations into multiple masked processing-in-memory operations
US 12,417,024↗
2025August 12, 2025
Method and apparatus for quantization and dequantization of neural-network data using processing-in-memory
US 12,387,767↗
2025July 8, 2025
Near-memory engine for reducing bandwidth utilization in sparse data applications
US 12,353,715↗
2025January 28, 2025
Compiler directed fine-grained power management
US 12,210,398↗
2025January 14, 2025
Memory sprinting
US 12,197,735↗
2024September 10, 2024
Memory sprinting
US 12,086,418↗
2023September 12, 2023
Method and apparatus for recovering regular access performance in fine-grained DRAM
US 11,756,606↗
2023August 15, 2023
Compiler directed fine-grained power management
US 11,726,546↗
02
International grants
32025Japan
Method and apparatus for recovering regular access performance in fine-grained DRAM
JP 7,595,229 B1↗
2025South Korea
Method and apparatus for recovering regular access performance in fine-grained DRAM
KR 10-2812912 B1↗
2025China
Method and apparatus for recovering regular access performance in fine-grained DRAM
CN 118525335 B↗
03
Published pending applications
42026Published June 25, 2026
Storing popularity metadata for embedding vectors
US 2026/0178554 A1↗
2026Published April 2, 2026
Method and apparatus for efficient memory setting
US 2026/0093616 A1↗
2024Published November 28, 2024
Logic die-based techniques for DRAM row segmentation and fine-grained accesses on stacked memory
US 2024/0395289 A1↗
2024Published January 4, 2024
Allocating memory for processing-in-memory devices
US 2024/0004786 A1↗