Extended Memory Microcode Architecture for Data Processing Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As the size and quantity of data stored in memory devices increase, transferring data to and from a host becomes time-consuming and resource-intensive, leading to increased processing time and resource consumption, especially when performing memory operations on large blocks of data.
Innovation Solution
The implementation of an extended memory architecture that allows for the performance of memory operations using a single address and operand, enabling computing devices to execute operations within the device without external commands, thereby reducing the need for locking or mutex operations and optimizing data transfer through multiple communication subsystems, such as PCIe and AXI interconnect interfaces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is transferred to and from host for memory operations, then data storage capability is improved, but processing time and resource consumption increase
Solution Approach 1:
The patent divides the memory system into compute devices with local memory arrays, separating computation from centralized host memory. Each compute device can perform operations on locally stored data without requiring transfers to the host, reducing processing time while maintaining storage capability.
Solution Approach 2:
The patent introduces compute devices as intermediary components between the host and memory arrays. These compute devices execute operations locally on memory data, acting as mediators that eliminate the need for continuous host-memory transfers and reduce processing time.
2Quantity of substance
If data is transferred to and from host for memory operations, then data storage capability is improved, but resource consumption increases
Solution Approach 1:
The patent segments the memory system into distributed compute devices with local memory arrays. This segmentation allows operations to be performed locally without transferring data to the host, reducing energy consumption while maintaining storage capability.
Solution Approach 2:
The compute devices perform memory operations autonomously using locally stored data, serving themselves without requiring host intervention. This self-service capability eliminates energy-consuming data transfers between host and memory while maintaining full storage functionality.
3Quantity of substance
If multiple commands and function calls are used for memory operations, then data storage capability is improved, but processing efficiency decreases
Solution Approach 1:
The patent combines multiple memory operations into single commands executed by compute devices. Instead of requiring separate function calls for each operation, the system merges read, write, and compute operations into unified commands that operate on locally stored data, improving processing efficiency while maintaining storage capability.
4Quantity of substance
If locking or mutex operations are implemented for memory operations, then data storage capability is improved, but device complexity increases
Solution Approach 1:
The patent segments memory access across multiple compute devices, allowing parallel operations without conflicts. This segmentation eliminates the need for locking mechanisms as each compute device operates independently on its local memory array, reducing device complexity while maintaining storage capability.
Data Source
AI summary
Systems, apparatuses, and methods related to extended memory microcode components for performing extended memory operations are described. An example apparatus can include a plurality of computing devices. Each of the computing devices can include a processing unit and a memory array. The example apparatus can include a plurality of microcode components coupled to each of the plurality of computing devices and each comprise a set of microcode instructions. The example apparatus can further include a communication subsystem coupled to a host and to each of the plurality of computing devices. Each of the plurality of computing devices can be configured to receive a request from the host, retrieve at least one of the set of microcode instructions, transfer a command and the at least one of the set of microcode instructions, and receive a result of performing the operation.