Patent Leaderboard
USPTO Patent Rankings Data through Dec 31, 2025
Home › Loss-scaling for deep neural network training with reduced precision

Loss-scaling for deep neural network training with reduced precision

US Patent 11842280 · Granted Dec 12, 2023

Estimated economic value: $2,068,008,000 ⓘ

Assignee

Inventors

View full patent text on Google Patents →

More Patents by These Inventors

Increasing sparcity in data sets2025
Generalized acceleration of matrix multiply accumulate operations2025
Techniques to power balance multiple chips2025
Voice assisted remote screen sharing2025
LiDAR and mobile device2025
Leveraging low power states for fault testing of processing cores at runtime2024
Text input synchronization for remote applications2024
Transposed sparse matrix multiply by dense matrix for neural network training2024
Telephone call information collection and retrieval2024
User-defined metered priority queues2024
Secure digital assistant integration in web pages2023
Generalized acceleration of matrix multiply accumulate operations2023