Patent Leaderboard
USPTO Patent Rankings Data through Dec 31, 2025
Home › Power savings for neural network architecture with zero activations during inference

Power savings for neural network architecture with zero activations during inference

US Patent 10817042 · Granted Oct 27, 2020

Estimated economic value: $34,955,000 ⓘ

Assignee

Inventors

View full patent text on Google Patents →

More Patents by These Inventors

Load store bank aware thread scheduling techniques2025
Multi-render partitioning2025
Multi-render partitioning2025
Graphics processing unit processing and caching improvements2025
Graphics processing unit processing and caching improvements2025
Immediate offset of load store and atomic instructions2025
Unified stateless compression system for universally consumable compression2025
Coordination and increased utilization of graphics processors during inference2025
Coordination and increased utilization of graphics processors during inference2025
Coordination and increased utilization of graphics processors during inference2025
Compression for sparse data structures utilizing mode search approximation2025
Dynamically scalable and partitioned copy engine2025