Cache Line Metadata Decryption Using Hash-Based Parallel Cryptography
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing encryption methods, such as AES-XTS, incur significant latency when encrypting and decrypting cache line metadata in modern computer systems, particularly in environments like CXL, which are sensitive to latency.
Innovation Solution
A cryptographic circuit that utilizes a hash function and a pseudorandom function to generate a bit sequence combined with cache line metadata for encryption/decryption, reducing the overall latency by performing these operations in parallel and requiring less than 17 cycles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional AES-XTS cryptographic methods are used to encrypt and decrypt cache line metadata along with data, then data security is improved, but decryption latency increases significantly (28 clock cycles)
Solution Approach 1:
The patent segments the cryptographic processing into two independent parts: data encryption/decryption using AES-XTS and metadata encryption/decryption using a separate hash-based pseudorandom function. This segmentation allows both operations to proceed in parallel without interfering with each other, reducing total latency while maintaining security for both data and metadata.
Solution Approach 2:
The patent performs preliminary action by pre-computing hash values of encrypted data during the data encryption phase. These hash values are then used during decryption to quickly generate the pseudorandom key for metadata decryption, eliminating the need for separate cryptographic operations and reducing overall decryption latency.
2Reliability
If cryptographic operations are performed on both cache line data and metadata, then security protection is improved, but processing speed deteriorates
Solution Approach 1:
The patent divides cryptographic processing into separate segments for data and metadata, allowing them to be processed independently and in parallel. The data portion uses AES-XTS while the metadata portion uses a hash-based pseudorandom function, enabling both to be secured without sequentially processing both, thus maintaining security while improving processing speed.
Solution Approach 2:
The patent changes the cryptographic parameter approach by using different cryptographic methods for data and metadata: AES-XTS for data and a hash-based pseudorandom function for metadata. This parameter differentiation optimizes processing speed for each type while maintaining appropriate security levels, improving overall productivity without compromising security.
Data Source
AI summary
Systems and techniques for cryptographically protecting data in a computer memory are disclosed. The techniques include dividing the data into a first portion and a second portion, encrypting the first portion of the data to create a first stored form of the data, encrypting the second portion of the data, and storing, in the computer memory, the first stored form of the data and a second stored form of the data. The techniques include, to encrypt the second portion, calculating a hash based on the first stored form of the data, applying a first pseudorandom function to the hash to obtain a bit sequence, and combining the bit sequence with the second portion of the data to obtain the second stored form of the data.