Agentic AI Reference Architecture

Infrastructure · Data artifact

Unpartitioned GPU Layout

Data artifactInfrastructureInfrastructurearc:UnpartitionedGPULayout

A GPU partition layout that disables hardware partitioning so each GPU is allocated whole to one workload.

Responsibility. Declares that cluster GPUs are exposed as full, dedicated devices.

Also known as: Strategy none

Variant of GPU Partition Layout abstract

When to choose. Choose for 70B+ models, high-concurrency single-tenant or latency-critical deployments needing a full GPU.

specializesconfiguresis target of alternativeTois target of alternativeToGPU Partition Layout: specializesGPU Partition LayoutDedicated GPU Device: configuresDedicated GPU DeviceMixed GPU Partition Layout: is target of alternativeToMixed GPU Partition LayoutUniform GPU Partition Layout: is target of alternativeToUniform GPU Partition La…
Direct neighbourhood (hover for relationship types)

Relationships

configures structural

alternative to variability

Sources

  1. Ch7.6: T. Nguyen, "Multi-Instance GPU," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 7.6. ISBN: 9798244538229.