Pii AI RUNTIME INFRASTRUCTURE

One brain to orchestrateOne cerebellum to execute

PiiEngine organizes compute and intelligence at the enterprise center; PiiHive executes real-time workloads at the edge; PiiOS and PiiIron provide the common software foundation and hardware platform

View architecture →Enterprise AI Engine ↓Edge AI Platform ↓

01ARCHITECTURE

1 FOUNDATION · 2 PLATFORMS · 1 HARDWARE SYSTEM

One common foundation,two computing platforms

PiiOS is the unified base. PiiEngine serves the enterprise center, PiiHive serves edge sites, and PiiIron provides tiered hardware products

APPLICATION AREASRail TransitEnergyCommercial SpaceEmbodied AI

01 / COMPUTING PLATFORMENTERPRISE CENTER

BRAIN

PiiEngine

Enterprise AI Engine

A unified platform for virtualization, cloud-native workloads, intelligent computing and agents

VMVirtualizationCCCloud NativeGMAI ComputeAGAgents

Models · Apps · Policies⇄Status · Data · Tasks

02 / COMPUTING PLATFORMEDGE SITES

CEREBELLUM

PiiHive

Edge AI Computing Platform

Edge scheduling, dual-kernel real-time operation, OTA management and local AI inference

ECEdge ComputeRTReal-time KernelsOTARemote EvolutionAILocal Inference

03 / FOUNDATIONHEART

PiiOS

Immutable Base OS

A secure, recoverable and evolvable base for both computing platforms

04 / HARDWARELIMBS

PiiIron

Hardware Portfolio

Enterprise, desktop and lightweight edge hardware products

ENTERPRISE CENTEREDGE SITEINTELLIGENT DEVICE

02ENTERPRISE AI ENGINE

PiiEngine / ENTERPRISE AI ENGINE

Unify diverse workloadsEvolve enterprise AI continuously

PiiEngine brings existing services, cloud-native platforms, AI computing and agents onto one operational foundation

ENTERPRISE AI ENGINE

PiiEngineEnterprise AI Engine

FROM TRADITIONAL IT TO AI PRODUCTION

Legacy continuity

Carry forward virtual machines and existing services through phased migration

Protect existing operations

Compute orchestration

Coordinate CPU, GPU, storage and network resources across workloads

Improve utilization

AI production

Bring model services, AI applications and agents into one governed runtime

Operationalize AI

Lifecycle operations

Unify deployment, monitoring, upgrade, rollback and lifecycle management

Reduce complexity

Open adaptation

Support heterogeneous compute and domestic hardware ecosystems

Adapt across environments

Continuous evolution

Move from traditional IT to cloud-native, AI compute and agents without rebuilding

Preserve investment

Virtualization→Cloud Native→AI Compute→Agents

03EDGE AI PLATFORM

PiiHive / EDGE AI COMPUTING PLATFORM

Bring intelligence to the edgeClose the execution loop on site

PiiHive brings models, applications and policies to edge sites for real-time operation, local inference, offline autonomy and remote evolution

EDGE AI COMPUTING PLATFORM

PiiHiveEdge AI Computing Platform

FROM CENTER COLLABORATION TO FIELD AUTONOMY

Edge scheduling

Coordinate nodes, applications and heterogeneous edge resources

Organize field resources

Real-time runtime

Support low-latency perception, inference and control workloads

Run time-sensitive tasks

Local inference

Process data and run AI models locally at the edge

Keep data on site

Offline autonomy

Keep critical workloads running when central connectivity is unavailable

Maintain continuity

OTA evolution

Update systems and applications remotely with recovery and rollback

Reduce upgrade risk

Unified control

Monitor nodes, distribute policies and manage versions centrally

Simplify edge operations

Central Delivery→Edge Runtime→Local Autonomy→Status Return

04FOUNDATION & HARDWARE

PiiOS & PiiIron / HEART & LIMBS

A common base and a tiered hardware portfolio

PiiOS supplies the operating-system foundation; PiiIron supplies enterprise, desktop and lightweight edge hardware

PLATFORM 01 / BRAINPiiEngineEnterprise AI Engine

PLATFORM 02 / CEREBELLUMPiiHiveEdge AI Computing Platform

UNIFIED INTERFACES · CONSISTENT RUNTIME · COORDINATED GOVERNANCE

FOUNDATION / HEART

PiiOSImmutable Base Operating System

UNIFIED RUNTIME FOUNDATION

IMMUTABLEImmutable systemArtifact-based delivery prevents configuration drift

REAL-TIMEReal-time runtimeExecution for time-sensitive and critical workloads

SECURITYBuilt-in securitySystem-level isolation and trusted boundaries

RECOVERYRecoverable updatesRemote upgrades, recovery and rollback

HETEROGENEOUSHardware adaptationSupport for diverse processors and compute platforms

ADAPT · OPTIMIZE · VERIFY · DELIVER

HARDWARE / LIMBS

PiiIronHardware Product Portfolio

PiiMAX Pro / MAXPiiMini Pro / MiniPiiGateway / Node

05FOCUS AREAS

SOLUTIONS

Apply the product system to mission environments

Explore dedicated solution paths for physical intelligence and space intelligence

SOLUTION / 01

Embodied AI

Central model services, real-time edge platforms and device execution

Explore →

SOLUTION / 02

Commercial Space

Ground governance, mission computing and on-orbit runtime

Explore →

TECHNICAL CONSULTATION

Turn product capabilities into operational outcomes

Define an implementable and evolvable path around existing systems, compute resources and deployment boundaries

Talk to our team ↗