Proximal Unified Compute

Leapfrog
into AI.

Don't build an AI stack. Deploy one. PUC is a complete, rack-ready appliance — hardware, software, models, and integrations — at half the cost of assembling it yourself.

Explore the appliance View configurations

Lower TCO

0min

Time to value

AI engineers needed

Made in India

Sovereign by design

Explore in 3D

Pull the appliance apart.

An interactive model of the Compute.AI node. Drag to rotate, scroll to zoom, hover any part to inspect it — or hover the list to highlight it on the model.

Drag · Scroll · Hover

Front bezel

Red status line · vents

x86 CPU

Agentic Compute · 24–128 cores

GPU

LLM · Tensor · Prefill

xPU

Custom · Decode

HBM memory

128 – 512 GB · at the center

800G UEC ports

4× uplinks · 3.2 Tbit / node

Hardware configurations

Three sizes. One architecture.

Memory-centric, air-cooled, rack-ready. Choose the configuration that fits your workload.

Entry

Small

SMBs, departments & branch offices

Up to 30 concurrent users

GPUHigh-performance GPU

GPU Memory64 – 96 GB

CPUMulti-core CPU

System RAM128 GB

Network10/25/100 GbE

Storage2× 4 TB NVMe

Ideal for remote offices, retail stores, and single-department deployments.

Professional

Medium

Multi-department & mid-market

Up to 100 concurrent users

GPUHigh-performance GPU

GPU Memory128 – 192 GB

CPUMulti-core CPU (dual)

System RAM256 GB

Network10/25/100 GbE

Storage2× 8 TB NVMe

Recommended for mid-market customers and small manufacturing plants.

Enterprise

Large

Enterprise & regulated industries

100+ concurrent users

GPUHigh-performance GPU

GPU Memory192 – 384 GB

CPUMulti-core CPU (dual)

System RAM256 GB

Network10/25/100 GbE

Storage10× 8 TB NVMe

Built for enterprise customers in regulated industries with high-throughput requirements.

All configurations include the full software stack: PAC engine · AIOps · Data Lake + Integrations · Inference Engine

Designed & built in India

3 tiers

Configurations

384 GB

Max GPU memory

80 TB

Max NVMe

100 GbE

Networking