Back to results
US 2026/0087411 A1Published application
Apparatus with expanded artificial intelligence training circuit and methods for operating the same
AssigneeMicron Technology, Inc.
InventorsRohit Sehgal, Nitin N. Okhade, Vishal Tanna, Rohit Sindhu
Life of this patent
Estimating the term…
Abstract
Methods, apparatuses, and systems related to an Artificial Intelligence (AI) training system are described. The training system includes a memory device external to a central processing module and configured to generate preprocessed training data from raw training data. The preprocessed training data may be applied to an AI model, thereby training the AI model using data preprocessed at the memory device.
Claims20
A computing system comprising:
- a central computing module configured as a master processor for the computing system, wherein the central computing module is configured to access raw training data;
- a memory module external to and communicatively coupled to the central computing module, the memory module configured to:
- receive the raw training data from the central computing module;
- and
- generate a preprocessed training data based on operating on the raw training data;
- and
- an accelerator module communicatively coupled to the central computing module and configured to perform computations as commanded by the master processor, wherein the performed computations include training an Artificial Intelligence (AI) model using the preprocessed training data.
4 more elements
A memory device comprising:
- a communication interface configured to communicate with a Central Processing Unit (CPU) for training an Artificial Intelligence (AI) model;
- a set of memory cells coupled to the communication interface and configured to store raw training data provided by the CPU;
- and
- local processors coupled to the set of memory cells and configured to generate preprocessed training data based on reformatting the raw training data, wherein the preprocessed training data is configured to be used to train the AI model.
1 more element
A method of operating a computing system, the method comprising:
- obtaining raw training data at a memory circuit external to a central processing circuit;
- generating preprocessed training data based on reformatting the raw training data using processors local to the memory circuit;
- and
- communicating the preprocessed training data to the central processing circuit, wherein the communicated preprocessed training data is configured to be applied to an Artificial Intelligence (AI) model for training the AI model.
1 more element
Dependencies are inferred from the claim text.