Skip to main content
Version: v2511

What is AIBooster?

AIBooster is a performance engineering platform for continuously observing and improving the performance of AI workloads.

  • PO: Performance Observability
    • 🔍 Visualization: View usage rates and efficiency of various hardware at a glance
    • 📊 Analysis: Identify software bottlenecks and discover improvement opportunities
  • PI: Performance Intelligence
    • ⚡ Performance Improvement: Continuously improve performance with automatic tuning
    • 💰 Cost Reduction: Reduce inefficient resource usage and improve ROI

Through visualization dashboards, users can visualize the utilization efficiency of various hardware resources such as CPU, GPU, interconnect, and storage, as well as software bottlenecks, to analyze the performance characteristics of AI workloads. Furthermore, by applying optimization frameworks designed for AI workloads, efficient performance improvements are possible.

Start fast and cost-effective AI training and inference with AIBooster!

Feature Highlights​

Inference Model Profiling​

AcuiRT is a framework that supports model conversion for inference. With this release, the ConversionWorkflow feature has been added, enabling easy performance profiling of converted models. Users can obtain the following information about converted models by adding just a few lines of code:

  • Model computational accuracy (Accuracy)
  • Model processing time (Latency)
  • Profiling results from PyTorch Profiler

For instructions on how to use ConversionWorkflow, please refer to this page.

Autonomous Tuning of Performance Parameters​

ZenithTune can automatically discover jobs that meet specific conditions in a Kubernetes environment and autonomously optimize the hyperparameters of these jobs. By enabling this feature, you can continuously search for performance-optimal parameters without manually running tuning jobs. In this release, new features have been added to customize tuning jobs in order to maximize the return on the cost required for tuning.

  • Job Filtering
    • Flexibly specify conditions for jobs to be tuned.
  • Job Routing
    • Apply different tuning configurations based on job type and purpose.
  • Custom Patches
    • Override tuning job settings from PyTorchJob manifests.

For instructions on how to use autonomous tuning, please refer to this page.

✨ Guides​

Quick Start Guide

Learn about AIBooster overview, setup methods, and basic usage.


Performance Observation Guide

Learn how to use visualization dashboards to observe AI workload performance.


Performance Improvement Guide

Learn how to use frameworks to improve AI workload performance.