Distributed Text-to-Speech Processing Cost Scheduling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Text-to-speech (TTS) processing systems face inefficiencies due to varying server costs and demand, making it challenging for cost-sensitive users to obtain affordable processing without compromising on time or quality.

Innovation Solution

A distributed TTS processing system that allows users to schedule requests on lower-cost servers during off-peak times, considering factors like cost, result turnaround time, and quality, by dividing requests into logical portions and allocating them across multiple servers for efficient processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If TTS requests are processed on high-demand servers during peak times, then processing speed and reliability are improved, but processing costs increase significantly

Engineering Contradiction:
Improveprocessing reliabilityVSAvoidprocessing cost
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The system dynamically adjusts server selection and request scheduling based on real-time server load, demand patterns, and cost metrics. The load balancer continuously monitors server conditions and redistributes requests to optimize the balance between processing reliability and cost efficiency, allowing the system to adapt to varying operational conditions rather than using a static allocation strategy

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes operational parameters such as server selection criteria, request batching size, and scheduling intervals based on current system state. By adjusting these parameters dynamically, the system can process requests on lower-cost servers during off-peak times while maintaining reliability through parameter optimization, resolving the contradiction between cost and reliability

Inventive Principle:
Principle #35Parameter changes

2Loss of energy

If TTS requests are processed on lower-cost servers during off-peak times, then processing costs are reduced, but processing time increases

Engineering Contradiction:
Improveprocessing costVSAvoidprocessing time
Core Design Contradiction:
Loss of energyVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-processing requests, pre-warming server resources, and advance scheduling of non-urgent requests during off-peak hours. By preparing data and resources in advance, the system can process requests on lower-cost servers without significant delays, as the preliminary work reduces the actual processing time required when execution occurs

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system maintains continuous useful action through background pre-processing, caching of intermediate results, and overlapping of request preparation with processing. This allows the system to utilize lower-cost servers during off-peak times while maintaining overall throughput, as the continuous action ensures that no time is lost to idle waiting periods

Inventive Principle:
Principle #20Continuity of useful action

3Loss of energy

If TTS requests are divided into multiple portions and processed across different servers, then cost efficiency is improved, but system complexity increases

Engineering Contradiction:
Improveprocessing costVSAvoidsystem complexity
Core Design Contradiction:
Loss of energyVSDevice complexity

Solution Approach 1:

The system segments TTS requests into logical portions that can be independently processed across multiple servers. Each segment is assigned to appropriate servers based on cost, capacity, and availability, allowing parallel processing that reduces overall cost. The segmentation is managed through automated routing rules that prevent excessive complexity by using clear, manageable divisions rather than arbitrary fragmentation

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system introduces intermediary components such as request brokers, result aggregators, and coordination services that manage the complexity of distributed processing. These intermediaries handle the coordination between server segments, manage data flow, and consolidate results, thereby reducing the complexity burden on individual servers and providing a simplified interface for users while enabling cost-efficient distributed processing

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS9311912B1Cost efficient distributed text-to-speech processing
Publication Date: 2016.04.12 AMAZON TECH INC
  • US9311912B1 patent drawing
  • US9311912B1 patent drawing
  • US9311912B1 patent drawing

AI summary

Text-to-speech (TTS) processing systems may be divided among remote TTS servers which are accessible through a network connection to local user devices. The costs for performing processing on these servers may vary according to time. To improve efficiency of TTS processing certain requests may be scheduled during low cost server times. A user may indicate a preference for such low cost delivery. A user may also indicate a preference for quick turnaround time, permitting scheduling of TTS processing during higher cost server times. A TTS processing system may also consider quality of TTS results when scheduling server processing time for a particular TTS request and may allocate more server time when higher quality results are desired.