Patent Leaderboard
USPTO Patent Rankings Data through Dec 31, 2025
Home › Compression of fully connected / recurrent layers of deep network(s) through enforcing spatial locality to weight matrices and effecting frequency compression

Compression of fully connected / recurrent layers of deep network(s) through enforcing spatial locality to weight matrices and effecting frequency compression

US Patent 11977974 · Granted May 7, 2024

Estimated economic value: $15,644,000 ⓘ

Assignee

Inventors

View full patent text on Google Patents →

More Patents by These Inventors

Interconnection structure and semiconductor package including the same2025
Two-dimensional mesh for compute-in-memory accelerator architecture2025
Four-bit training for machine learning2025
Padding input data for artificial intelligence accelerators2025
Organic light-emitting diode display with passive leakage-reducing structures2025
Stochastic rounding switched capacitor computation cores useful for efficient de2025
Split pulse width modulation to reduce crossbar array integration time2025
Single function to perform combined matrix multiplication and bias add operation2025
Neural network circuitry having floating point format with asymmetric range2025
Neural network circuitry having floating point format with asymmetric range2025
Neural network circuitry having floating point format with asymmetric range2025
Preemption in a machine learning hardware accelerator2025