SHA Hashing via SIMD Parallel Message Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Secure Hash Algorithm (SHA) implementations are time-consuming due to multiple shifts, XOR, and ADD operations, limiting their applicability in high-speed secure network applications and dragging server performance.
Innovation Solution
The use of Single Instruction Multiple Data (SIMD) instructions to perform SHA message scheduling in parallel, generating hash values for multiple message blocks simultaneously and compressing them serially, thereby reducing computation time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional sequential SHA computation methods are used, then implementation simplicity is maintained, but computation time increases substantially
Solution Approach 1:
The message schedule is divided into multiple segments that can be processed in parallel. The computation is segmented into independent chunks that can be handled by different processing units simultaneously, reducing overall computation time while maintaining manageable complexity through modular organization
Solution Approach 2:
The patent transitions from sequential single-threaded computation to parallel multi-threaded computation, adding a temporal dimension to the processing. Multiple message schedules are computed simultaneously across different threads rather than one after another, effectively utilizing another dimension of computation space
2Productivity
If multiple SHA operations are performed sequentially, then resource requirements are reduced, but throughput decreases
Solution Approach 1:
Multiple SHA operations are merged into a single parallel processing framework. The patent combines multiple message schedule computations into one coordinated parallel execution, allowing simultaneous processing of multiple operations while optimizing resource utilization across the processor cores
Solution Approach 2:
The patent performs preliminary setup of parallel processing environments, pre-configuring multiple threads and memory structures before execution. This preliminary action enables efficient parallel computation by having all processing units ready and positioned to execute simultaneously, maximizing throughput from the start
Data Source
AI summary
A processing apparatus comprises logic to, according to a selected secure hash algorithm (SHA) algorithm, generate hash values by preparing message schedules for a plurality of message blocks in parallel using single instruction multiple date (SIMD) instructions and performing compression in serial, and logic to generate a message digest conforming to the secure hash algorithm (SHA) algorithm.


