Fixed-Point ML Computing Architecture for Faster Parallel Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
General purpose processors are inefficient in processing large loads, leading to delays in machine learning tasks such as training artificial intelligence models, due to their limited operating speed.
Innovation Solution
A computation device and method that includes a controller unit, operation unit, and conversion unit, utilizing a primary processing circuit and secondary processing circuits to perform machine learning computations by converting input data into fixed-point data and executing operations efficiently, with a DMA unit and storage unit for data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If general purpose processors are used to process large loads, then versatility is maintained, but processing speed and efficiency deteriorate
Solution Approach 1:
The processor is segmented into a control unit and multiple operation units with specialized functions. The control unit handles instruction decoding and coordination, while operation units perform specific computational tasks in parallel, thereby maintaining versatility through modular design while improving processing speed through specialized parallel execution
Solution Approach 2:
The system changes the operational parameters by switching between different data formats (floating-point and fixed-point) and using variable precision arithmetic. This allows the processor to optimize for speed when using fixed-point arithmetic while maintaining the ability to use floating-point when precision is required, thus balancing versatility and processing speed
2Productivity
If general purpose processors operate at high speed, then processing efficiency improves, but power consumption increases
Solution Approach 1:
The system uses variable precision arithmetic where full floating-point precision is applied only when necessary, while fixed-point arithmetic with lower precision is used for operations where approximate results suffice. This partial application of high-precision processing improves efficiency while reducing the power consumption associated with maintaining full precision across all operations
Solution Approach 2:
The processor dynamically changes operational parameters including data format (floating-point vs. fixed-point) and precision level based on the computational requirements. This parameter adaptation allows the system to achieve high processing efficiency when needed while consuming less power during operations that can tolerate lower precision
Data Source
AI summary
The present disclosure provides a computation device. The computation device is configured to perform a machine learning computation, and includes an operation unit, a controller unit, and a conversion unit. The storage unit is configured to obtain input data and a computation instruction. The controller unit is configured to extract and parse the computation instruction from the storage unit to obtain one or more operation instructions, and to send the one or more operation instructions and the input data to the operation unit. The operation unit is configured to perform operations on the input data according to one or more operation instructions to obtain a computation result of the computation instruction. In the examples of the present disclosure, the input data involved in machine learning computations is represented by fixed-point data, thereby improving the processing speed and efficiency of training operations.