Abstract: A system for optimizing the use of append-only strand data structures is disclosed, with a device driver that transmits strand commands to firmware on a storage device. The storage device firmware executes strand commands natively on the storage device without needing to transmit data over the system bus to copy data to/from a strand saved on the storage device.

Abstract: A system for providing high availability to persistent data is provided that persists data on a node in discrete partitions that are each backed up on separate backup systems. When the system detects that the data saved on the node is compromised, the system could restore the data from multiple backup systems. Since the backup systems hold discrete sets of data that do not overlap one another, the backup systems are able to simultaneously and locklessly restore data to the compromised node.

Abstract: A persistent distributed database system is disclosed that partitions a database structure into a plurality of partitions to optimize repeated analytics that are executed on the database structure. A repeated analytic is partitioned into sub-analytics that are executed on the database partitions, and upon repeated executions, sub-analytics are only repeated upon database partitions that have been updated since the previous execution. This drastically reduces the time to repeat execution of a repeated analytic, and frees up resources that could be better spent on other processes.

Abstract: An improved persistent memory process is disclosed that persists a traversable data structure by converting the traversable data structure into one or more traversable array structures that are saved on a secondary memory. The traversable array structure could be rapidly traversed and maintained, and can be used to restore one or more portions of the traversable data structure.

Abstract: A system for providing both low-level, physical data access and high-level, logical data access to a single process is disclosed, having a data block table with a physical memory address portion and a logical memory address portion. Data blocks that are mapped to physical memory bypass multiple logical memory address layers, such as the operating system layer and a logical block address layer, while data blocks that are mapped to the logical memory will be routed through traditional API layers, providing both increased performance and flexibility.

Abstract: An improved transmission system that utilizes a shared buffer accessible by multiple threads, processes, or other discrete systems of different producing speeds is disclosed which limits transmission bottlenecks which occur when producers write data at different speeds. The system reserves portions of the shared buffer for each of the different producers, and allows the producers to write to the shared buffer asynchronously—even if the shared buffer is read from serially. This allows a fast producer to write its data to the shared buffer without needing to wait “in line” for slower producers.

Abstract: An improved search engine reorganizes an input database containing a set of values into a searchable tree structure having nodes of information with pointers to child nodes. The searchable tee structure is then converted into a searchable array structure that uses mathematical operations to rapidly traverse through elements of the array using mathematical formula to reach different memory locations instead of requiring pointer lookups.