Permission Control via Data Redundancy in Deterministic Streaming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cloud computing environments, shared computing resources face data security concerns and data leakage among tenants, particularly with modern AI chips that have significant memory resources and residual data that need protection. Clearing memory is time and power intensive, making it impractical for applications requiring fast compute times.
Innovation Solution
A deterministic streaming system with a tensor streaming processor (TSP) architecture uses data redundancy for permission control. This system encodes data vectors with permission control codes, updates these codes for different users, and employs error correction to ensure that data is accessible only to authorized users, preventing unauthorized access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If memory is cleared by systematically writing over contents, then data security is improved, but time consumption and power consumption increase significantly
Solution Approach 1:
The patent applies preliminary action by pre-processing data before storage through encoding transformations (such as bit inversion based on permission codes). This preprocessing ensures that when data needs to be secured, the encoding structure already exists to enable rapid security enforcement without requiring time-consuming systematic writing operations. The data is prepared in advance with embedded permission control mechanisms that can be activated instantly.
Solution Approach 2:
The patent changes the parameter of data representation by transforming data into encoded forms (e.g., inverting specific bits based on permission control codes). This parameter transformation allows the same physical memory location to represent different data states for different users, enabling security enforcement through parameter interpretation rather than physical data overwriting, thus reducing time and power consumption.
2Reliability
If memory is cleared by systematically writing over contents, then data security is improved, but power consumption increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-processing data before storage through encoding transformations (such as bit inversion based on permission codes). This preprocessing ensures that when data needs to be secured, the encoding structure already exists to enable rapid security enforcement without requiring time-consuming systematic writing operations. The data is prepared in advance with embedded permission control mechanisms that can be activated instantly.
Solution Approach 2:
The patent changes the parameter of data representation by transforming data into encoded forms (e.g., inverting specific bits based on permission control codes). This parameter transformation allows the same physical memory location to represent different data states for different users, enabling security enforcement through parameter interpretation rather than physical data overwriting, thus reducing time and power consumption.
3Reliability
If data is encoded with permission control codes for each user, then data security is improved, but device complexity increases
Solution Approach 1:
The patent applies local quality by implementing permission control at the bit level rather than requiring complex system-wide security mechanisms. Each data element can have specific bits inverted based on user-specific permission codes, creating localized security transformations that are simple to implement and manage. This approach distributes security functionality across individual data elements rather than requiring centralized complex security hardware.
Solution Approach 2:
The patent changes the parameter of data representation by transforming data into encoded forms (e.g., inverting specific bits based on permission control codes). This parameter transformation allows the same physical memory location to represent different data states for different users, enabling security enforcement through parameter interpretation rather than physical data overwriting, thus reducing time and power consumption.
4Reliability
If error correction codes are used to detect unauthorized access, then data security is improved, but manufacturing complexity increases
Solution Approach 1:
The patent applies universality by designing permission control codes that serve dual functions: they act as both access control mechanisms and error correction codes. The same encoding structure that provides user-specific data transformation also enables detection of unauthorized access attempts through error detection capabilities. This multi-functionality eliminates the need for separate security hardware and simplifies manufacturing by using a unified code structure for both purposes.
Solution Approach 2:
The patent changes the parameter of data representation by transforming data into encoded forms (e.g., inverting specific bits based on permission control codes). This parameter transformation allows the same physical memory location to represent different data states for different users, enabling security enforcement through parameter interpretation rather than physical data overwriting, thus reducing time and power consumption.
Data Source
AI summary
Embodiments are directed to a computing system with permission control via data redundancy. The computing system includes a memory and a permission control circuit coupled to the memory. The permission control circuit encodes a first data vector by using a bit position register with a first permission control code for a first user, writes the encoded first data vector into the memory, and updates content of the bit position register from the first permission control code to a second permission control code for a second user. The encoded first data vector written into the memory is inaccessible for the second user based on the updated content of the bit position register.