Program Initiation/Switching TRIZ Cases
- Bunch Buffer Load Latency Amelioration in Compute Arrays
Bunch buffers decouple control word loading from compute element execution cycles to enable autonomous operation within a two-dimensional array.
- Declarative Plugins for Cloud Batch Job Control
Declarative plugins interpret batch job definitions to select and process database records, eliminating compilation overhead in cloud environments.
- Multi-OS Runtime Virtualization for Cross-Platform Application Execution
A multi-OS runtime virtualizes guest environments to execute non-native applications on a host device.
- Backfill Scheduler Frequency Control for HPC Peak Power Reduction
Centralized parameter changes lower peak power usage by smoothing consumption without delaying job completion times.
- Automated Software Application Bundling for License Tracking
Automated software application bundling detects installed scanners and retrieves configuration files to execute licensed applications across computing devices.
- Virtual Machine Power Management via Dynamic Freezing
Dynamic freezing of idle virtual machines conserves processor cycles while maintaining system responsiveness.
- Dynamic User Interface App Scheduling via AI Policy
A dynamic user interface system filters installed applications using an artificial intelligence service to schedule availability based on predefined policies.
- Android Process Shutdown Prevention via Event Object Management
A process management apparatus removes identifiers from system event objects to prevent unintentional reinitiation of shut down processes.
- Stream Computing Checkpointing via Operator Merging
Merging shared stream computing data into single checkpoint operations reduces bandwidth usage and storage volume while maintaining data integrity.
- Task Orchestration via Auxiliary Container File System Sharing
Decoupling resource management from task execution enables hot updates and data delivery, resolving Kubernetes coupling limitations.
- Multi-Engine Protocol Architecture for Scalable Command Processing
A multi-engine protocol architecture segments command processing into parallel pipeline stages to boost throughput.
- Programmable Interface Unit for Parallel Data Transfer
A programmable interface unit separates software routines into categories to execute data transfer at distinct cycle rates.
- Live Migration Agent Configuration Transformation
Transforming static configuration files enables dynamic routing and zero-downtime transitions for cloud-native services.
- Accelerator Scheduler for Multi-Application Priority Access
Hardware scheduler prioritizes accelerator requests using embedded descriptors to optimize resource allocation across shared computing systems.
- Stream Processing Dispatch Adapting Parallelism to Thread Availability
A stream processing dispatch mechanism dynamically adjusts parallelism based on available physical threads.
- Neural Network Power Profile Estimation for AI Accelerators
Estimates neural network power profiles on a per-layer basis using hardware efficiency factors to resolve inadequate fine-grained profiling capabilities.
- Late-Binding Schema for Real-Time Anomaly Detection in Data Intake Systems
A data intake system uses a late-binding schema to process raw machine data and detect anomalies during ingestion.
- Multi-Flat-Map Publisher for Microservices Event Stream Flattening
A multi-flat-map publisher component concurrently flattens events from multiple publishers into a single stream for downstream subscribers.