OneShot Neural Architecture and Hardware Architecture Search
Summary
USPTO granted Google LLC patent US12596909B2 covering methods for jointly searching machine learning model architectures and hardware architectures in a combined search space. The technology utilizes weight sharing through a supernetwork and multi-objective reward functions optimizing for quality, performance, power, and area. The patent lists six inventors including researchers from Google's hardware and ML teams.
What changed
USPTO granted patent US12596909B2 to Google LLC for joint optimization of neural network architectures and hardware configurations. The invention enables simultaneous evaluation of models, hardware, and mapping strategies through weight sharing in a supernetwork, with a multi-objective reward function balancing quality, performance, power, and area constraints.
For organizations developing AI/ML hardware co-design tools or neural architecture search technologies, this represents proprietary prior art that may affect patent filing strategies and freedom-to-operate analyses. Technology companies researching similar joint optimization approaches for ML systems should consider potential licensing implications or design-around requirements.
What to do next
- Monitor for updates
Archived snapshot
Apr 7, 2026GovPing captured this document from the original source. If the source has since changed or been removed, this is the text as it existed at that time.
OneShot neural architecture and hardware architecture search
Grant US12596909B2 Kind: B2 Apr 07, 2026
Assignee
Google LLC
Inventors
Sheng Li, Norman Paul Jouppi, Garrett Axel Andersen, Quoc V. Le, Liqun Cheng, Parthasarathy Ranganathan
Abstract
Aspects of the disclosure are directed to jointly searching machine learning model architectures and hardware architectures in a combined space of models, hardware, and mapping strategies. A search strategy is utilized where all models, hardware, and mappings are evaluated together at once via weight sharing and a supernetwork. A multi-objective reward function is utilized with objectives for quality, performance, power, and area.
CPC Classifications
G06N 3/045 G06N 3/063
Filing Date
2022-07-28
Application No.
17875594
Claims
17
Related changes
Get daily alerts for USPTO Patent Grants - AI & Computing (G06N)
Daily digest delivered to your inbox.
Free. Unsubscribe anytime.
Source
About this page
Every important government, regulator, and court update from around the world. One place. Real-time. Free. Our mission
Source document text, dates, docket IDs, and authority are extracted directly from USPTO.
The summary, classification, recommended actions, deadlines, and penalty information are AI-generated from the original text and may contain errors. Always verify against the source document.
Classification
Who this affects
Taxonomy
Browse Categories
Get alerts for this source
We'll email you when USPTO Patent Grants - AI & Computing (G06N) publishes new changes.
Subscribed!
Optional. Filters your digest to exactly the updates that matter to you.