Success Story

As AI systems scale, network performance becomes critical to maximizing accelerated compute investments. At Nokia’s AI Networking Innovation Lab, Nokia and Keysight recreate realistic, multivendor AI environments to validate the complete technology stack. By testing synchronized workloads, congestion management, and emerging network technologies from design through deployment, the companies are turning AI networking from theory into proof and helping customers deploy infrastructure with greater confidence.

Nokia’s AI Networking Innovation Lab

Moving AI Networking from Theory to Proof

Case Study Highlights: Nokia’s AI Networking Innovation Lab

Defining the Next Generation of AI Networks

Nokia and Keysight bring together complementary expertise in AI data center networking and system-level validation. Nokia provides high-performance switching, network automation, and AI fabric innovation. Keysight provides workload emulation and benchmarking through AI Data Center Builder and AresONE hardware platforms. Working alongside GPU, server, NIC, storage, and optics partners, the companies recreate complete AI environments to determine how the entire technology stack performs together.

Validating AI Networks as Complete Systems

AI workloads turn the network into an integral part of a distributed computing system. Massive synchronized traffic bursts, collective operations such as AllReduce and AlltoAll, and congestion can cause an entire job to progress at the speed of its slowest flow. Metrics such as job completion time, tail latency, and bandwidth utilization cannot be validated one component at a time. Nokia needed to test switches, NICs, optics, servers, storage, software, and automation as one end-to-end system.

Realistic End-to-End AI Fabric Validation

At Nokia’s AI Networking Innovation Lab, Nokia and Keysight use Keysight AI Data Center Builder and AresONE platforms to emulate stateful AI training and inference workloads at scale. Testing with Nokia’s 7220 and 7250 data center switches covers Ultra Ethernet transport, packet trimming, congestion signaling, Micro-SID forwarding, and collective communication patterns. The lab also evaluates multivendor interoperability and automation across the lifecycle, from early design and silicon development through deployment.

From Network Design to Deployment Confidence

The collaboration gives Nokia a repeatable way to uncover integration issues earlier, benchmark system-wide performance, and create Nokia Validated Designs based on realistic AI workloads. By validating low-latency, lossless behavior and congestion management before deployment, Nokia can help customers reduce risk and improve network efficiency—supporting better GPU utilization, lower power consumption, shorter job completion times, and faster time to value.
James Wynia, Dell Technologies

"Together, Nokia and Keysight are helping the industry move AI networking from theory to proof so our customers can deploy their AI infrastructure knowing it will perform, not hoping it will."

Rudy Hoebeke, Nokia

Fireside Chat: Advancing AI Network Readiness

Experts from Nokia and Keysight discuss why AI networks must be validated as end-to-end systems rather than collections of isolated components. The conversation explores Ultra Ethernet, Multi-Path Reliable Connection, packet trimming, congestion signaling, Micro-SID forwarding, realistic workload emulation, and automation. It also offers practical guidance for designing, testing, and deploying low-latency, lossless AI fabrics.

AresONE-M 800GE

Validate equipment interoperability and performance.

Keysight AI Data Center Builder

Design and deploy AI network infrastructure​ and emulate AI workloads.

Explore Related Resources

White Paper

White Paper

7 Lessons for Designing and Validating AI Networks at Scale

Explore key considerations for designing, connecting, and validating AI network fabrics, from traffic behavior and topology choices to cabling, optics, load balancing, and full-system emulation at 800G, 1.6T, and beyond.

White Paper

White Paper

Scaling AI Data Centers for Real-World Performance

Technical insight on emulation, validation, and cluster behavior at AI infrastructure scale.

Report

Report

AI Cluster Networking Bottlenecks, Scale, and What Comes Next

Survey-based insights on infrastructure investment, emulation gaps, interconnect adoption, and optimization priorities in AI data centers.

×

Please have a salesperson contact me.

*Indicates required field

Select a preferred method of communication*Required Field
Preferred method of communication? Change email?
Preferred method of communication?

By continuing, you are providing Keysight with your personal data. See the Keysight Privacy Statement for information on how we use this data.

Thank you

A sales representative will contact you soon.