In-Die Arithmetic Logic Unit for Non-Volatile Memory Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for processing large datasets face significant performance issues due to slow DRAM access times and high costs associated with large amounts of volatile memory, making it inefficient and commercially unfeasible to handle datasets exceeding ten terabytes.
Innovation Solution
The implementation of a solid state drive unit with a bulk array of non-volatile memory cells and an arithmetic logic unit on an integrated circuit die, coupled with a dataset controller that performs operations locally without transferring data over the bus interface, enabling efficient processing of large datasets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If datasets are loaded into DRAM for processing, then processing speed is improved, but memory cost and system complexity increase significantly
Solution Approach 1:
The patent segments the dataset into multiple shards distributed across multiple storage devices. Each shard can be independently processed, allowing parallel processing without requiring the entire dataset to be loaded into DRAM simultaneously. This reduces memory requirements and system complexity while maintaining processing speed through concurrent operations on multiple shards.
2Productivity
If large amounts of DRAM are used to store datasets, then processing efficiency is improved, but cost increases making it commercially infeasible
Solution Approach 1:
The patent introduces a new dimension of processing by performing computations directly on storage devices (SSDs) rather than requiring data movement to DRAM. The SSDs execute processing operations locally on stored datasets, eliminating the memory bottleneck and enabling efficient processing of large datasets without requiring proportional DRAM capacity.
3Ease of operation
If data is transferred between DRAM and processor, then processing is enabled, but bandwidth limitations cause performance degradation
Solution Approach 1:
The patent extracts the processing capability from the traditional CPU-DRAM architecture and relocates it to the storage devices themselves. By embedding processing units within SSDs, the system eliminates the need for continuous data transfer between storage and processor, allowing datasets to remain stationary on SSDs while computations are performed in-place.
Data Source
AI summary
An apparatus comprises a bulk array of non-volatile memory cells on an integrated circuit die and an arithmetic logic unit on the die coupled to the bulk array. The arithmetic logic unit is operable to perform arithmetic logic operations on contents of the bulk array responsive to instructions received from outside of the die. The non-volatile memory cells may include NAND-type flash memory cells.