(KGNN) acts as a powerful bridge between (p 10 - p 11)

 

The Equitus Knowledge Graph Neural Network (KGNN) acts as a powerful bridge between IBM Power 10 and Power 11 by modernizing the data and AI layer on the existing Power 10 system, which then positions the workload for a performance-optimized migration to Power 11.

Here is how Equitus KGNN creates this seamless transition bridge:

__________________________________________________________________________

1. Modernizing the Workload on Power 10 (The Staging Area)

The Equitus KGNN platform delivers critical modernization benefits on the Power 10 system:

• AI-Ready Data Transformation: KGNN automatically unifies and transforms disparate structured and unstructured data sources into a single, scalable knowledge graph. This semantic layer is inherently AI-Ready, providing the context, explainability, and traceability crucial for modern AI and analytical models.

• Leveraging Power 10 Accelerators: KGNN is explicitly optimized to run natively on IBM Power 10 servers, directly utilizing the Matrix Math Accelerators (MMA) built into the Power 10 cores. This optimization allows users to deploy high-performance AI inferencing and real-time deep learning at the edge and on-premises without reliance on expensive GPUs or the cloud.

• On-Premises, Secure Foundation: By running KGNN on Power 10, users prove the value of their AI workloads in a private, secure, and fully controlled on-premises environment, which aligns perfectly with the Power platform's mission-critical reputation.

2. Creating the Seamless Bridge to Power 11 (The Accelerant)

The modernization effort on Power 10 directly facilitates the move to Power 11:

• Maximized Performance Uplift: Since KGNN workloads are already optimized for the Power architecture (specifically the MMA technology in Power 10), they are primed to take full advantage of Power 11's superior core performance (up to 45% more capacity than Power 10) and enhanced AI capabilities. The move from Power 10 to Power 11 is not just a hardware replacement; it is a direct performance multiplier for the already-tuned KGNN applications.

• Future-Proofing AI Investments: Power 11 is designed as an AI-ready infrastructure, featuring even more advanced accelerators and support for external accelerators like the IBM Spyre Accelerator. Moving the KGNN applications ensures the investment in the knowledge graph layer is preserved and future-proofed against the latest IBM hardware and software innovations (like Red Hat OpenShift AI integration).

• Enabling 'Right-Click and Move' Migration: IBM positions the Power 10 to Power 11 migration as a near-zero downtime event using a "right-click and move" approach for applications. Because the KGNN platform abstracts the application from the raw infrastructure by providing a unified data/AI layer, the physical hardware migration becomes simpler. The application logic running on the KGNN layer can be easily transitioned to the Power 11 environment, immediately benefiting from the generational performance boost with minimal disruption.

In summary, Equitus $\text{KGNN}$ allows Power 10 users to start their AI journey, establish mission-critical, high-performance knowledge graph workloads, and then perform a simple, high-value lift-and-shift of that modern workload onto the dramatically more powerful and resilient Power 11 infrastructure.