High-Performance 5G SA UPF

Intelligent. Cloud-Native. Carrier-Grade Resilience

Download Brochure

Ultra-Fast 5G: 750 Gbps, Lower CPU, Maximum Availability

Amantya’s 3GPP-compliant 5G SA UPF is a high-performance, cloud-native solution delivering ultra-low latency, massive throughput, and intelligent traffic management for core, edge, and far-edge networks

High Performance

Up to 750 Gbps per node with SmartNIC acceleration

Realistic Simulation

Sub-second failover and optimized data handling

Seamless Integration

Linear scalability with carrier-grade reliability

Actionable Insights

Core, edge, and far-edge ready with multi-interface support.

Actionable Insights

Dynamic optimization for consistent QoS

Actionable Insights

Fully 3GPP-compliant for simplified, standards-based rollouts

Get in Touch

Real World Outcomes

Test Cycle Acceleration  

Scales linearly for tier-1 traffic loads

Lower CPU Usage with SmartNIC Offload

>Power-efficient scaling, reduced TCO

Exceptionally reliable performance  

Carrier-grade reliability with 1s sync

Zero-touch onboarding 

Rapid deployment and Day-2 agility 

Key Differentiators 

Carrier-Grade Reliability

99.999% uptime with active-active HA and 1s session sync.

Zero-Touch Automation

YANG/gNMI, Prometheus, Grafana, and Helm/CI-CD integration for lifecycle management.

Seamless Hardware Support

Compatible with Intel and Mellanox cards for seamless deployment.

SmartNIC Acceleration

Napatech offload delivers 35% lower CPU load and reduced latency.

AI-Driven Traffic Tuning

Slice-aware QoS and predictive congestion control for consistent performance.

Cloud-Native Portability

Runs seamlessly on Kubernetes, Docker, VMs, or bare metal - one UPF image for all.

Solution Highlights 

AI-First Technolog. AI-Powered Workforce.

High-Performance Processing

Line-rate throughput up to 750 Gbps with DPDK/VPP acceleration for GTP-U, QoS, and traffic steering.

AI-First Technolog. AI-Powered Workforce.

Multi-Interface & Interworking

N3/N4/N6/N9 support plus integrated UPF, PGW-U, SGW-U for seamless 4G/5G fallback.

AI-First Technolog. AI-Powered Workforce.

Flexible Deployment Models

Runs in central core, edge DC, far-edge CPE, or hybrid multi-site clusters.

AI-First Technolog. AI-Powered Workforce.

Linear COTS Scaling

Scale by adding NICs - no proprietary ASICs. One binary image across all platforms. 

AI-First Technolog. AI-Powered Workforce.

Secure & Future-Ready Networking

Dual-stack IPv4/6 with zero-trust readiness and built-in 3GPP Release 17 compliance.

AI-First Technolog. AI-Powered Workforce.

Carrier-Grade Security

Lawful intercept (ETSI TS 101 671), DSCP marking.

AI-First Technolog. AI-Powered Workforce.

Active-Active High Availability

High uptime with sub-second session state sync for uninterrupted service.

AI-First Technolog. AI-Powered Workforce.

Built-In Observability

YANG/gNMI, Prometheus, and Grafana integration for real-time monitoring and analytics.

Deployment Flexibility

AI-First Technolog. AI-Powered Workforce.

Central Core

High-capacity UPF nodes (up to 750 Gbps) aggregating traffic from multiple regional sites. 

AI-First Technolog. AI-Powered Workforce.

Edge DC

Cloud-native UPF for metro or regional deployments with high session churn. 

AI-First Technolog. AI-Powered Workforce.

Far-Edge/CPE

Compact 1U or uCPE UPF for private 5G, air-gapped, and zero-trust environments. 

AI-First Technolog. AI-Powered Workforce.

Hybrid Clusters

Distributed UPF across sites with PFCP steering, multi-site HA, and centralized observability. 

Why Amantya UPF? 

AI-First Technolog. AI-Powered Workforce.

Unmatched Performance

Outperforms typical 200–500 Gbps vendor nodes with up to 750 Gbps per node.

Efficient Scaling

SmartNIC offload delivers CPU savings and reduces power use.

Intelligent QoS

AI-driven traffic tuning ensures slice-aware QoS and predictive congestion control.

Carrier-Grade Reliability

99.999% uptime with active-active HA and ultra-fast failover.

Cloud-Native & COTS-Ready

Deploy anywhere — core to far-edge — scale linearly and cut TCO.

Book a Demo

Have an idea in mind?

Let's Talk

4 + 4