Skip to content
Performance Intelligence

Intelligence in Every Path.

Patented machine-learning-based routing continuously evaluates available paths using permitted network and application telemetry to improve real-time application delivery conditions.

Request a Performance AssessmentDesign an AI Performance Pilot

Decision cycle

A five-stage decision cycle.

Performance Intelligence runs continuously across the Federation for each real-time workload. As conditions change, FGN can reevaluate available paths and select a different path based on the workload's current needs.

  1. Observe

    Continuously measure candidate paths and application demand.

  2. Evaluate

    Compare candidate paths against the application's qualitative requirements.

  3. Select

    Choose the path selected under current conditions for the application context.

  4. Deliver

    Serve the session over the selected path.

  5. Learn

    Feed observed outcomes back into the next evaluation.

See it by application

How selection changes with the workload.

Different applications value different network attributes. Choose an application to see which attribute Performance Intelligence prioritizes.

Application requirements

  • Fast, consistent response to interactive input.
  • Predictable behavior during live tournaments.
  • Support for local and regional participation.

Esports and Gaming sessions are judged by response, not just capacity. As conditions change, FGN can reevaluate available paths and select a different path based on the workload's current needs.

Candidate paths

  • Path A – shared consumer routeVariable
  • Path B – federation-preferred routeSelected
  • Path C – long-haul fallbackLonger path

Selected path. Path B is currently selected because its observed conditions match the responsiveness this session requires. Selection is continuously re-evaluated.

Want to follow the decision step by step? Open the route selection demo.

How path selection behaves

Path selection is continuous and contextual.

As conditions change, FGN can reevaluate available paths and select a different path based on the workload's current needs.

  • Network Conditions

    Latency, packet loss, jitter, congestion and route quality.

  • Application Context

    The type of real-time experience, policy and delivery requirement.

  • Compute Proximity

    Distance and path conditions between the user, edge and compute location.

Application examples

One intelligence capability. Three real-time Solutions.

The same path-selection foundation supports different application and business requirements.

How performance is measured

The context around every measurement.

Performance is reported in the context that produced it: the application, the users, the destinations, the baseline and the time window. Program-specific results are shared with each customer inside their pilot.

  • Application context

    What was being delivered

    Every measurement is tied to a specific real-time application (gaming session, inference call or simulation task) so results reflect the workload that was actually served.

  • Network context

    Where and when

    Latency, packet loss, jitter and route conditions are reported against the participating footprint and the time window observed in the pilot.

  • Baseline and evidence

    What it is compared to

    Improvements are described against the customer's own pre-pilot baseline. Numbers are shared inside the program rather than published as universal figures.

Caption: Performance figures are reported per program, in the context that produced them, and shared directly with each customer inside their pilot.

Why it matters commercially

The business value of continuous path selection.

Performance Intelligence turns real-time responsiveness into commercial outcomes for four groups of stakeholders.

  • For customers and end users

    More responsive sessions in the moment, so gaming, AI and simulation applications feel usable and predictable.

  • For operators

    A performance story that differentiates the network and supports new services on top of existing infrastructure.

  • For application and AI partners

    A path to place real-time workloads closer to demand, with measurable results inside each pilot.

  • For programs and sponsors

    Consistent, evidence-backed experience for cohorts across geography, so program outcomes travel with the participant.

Explore the differentiation

Connect the technical story to a business objective.

A performance assessment starts with the application, users, destinations, current baseline and the evidence needed to evaluate a pilot.

Request a Performance AssessmentDesign an AI Performance Pilot

Frequently asked

Straight answers.

  • Does FGN guarantee lower latency or packet loss?

    No. FGN improves real-time application delivery conditions through intelligent path selection. Specific latency and packet-loss improvements are reported for a defined pilot with an established baseline rather than as a universal number.

  • What information does Performance Intelligence consider?

    The approved public description includes factors such as latency, packet loss, jitter, route quality and proximity to compute. Data access remains subject to permitted telemetry and operating agreements.

  • Is Performance Intelligence a separate market-facing pillar?

    No. It is a differentiating capability beneath the Federated Gig Network and appears on deeper technical and sales-diligence pages.