Keysight Unveils AI Data Center Builder for Optimizing AI Infrastructure

Keysight Unveils AI Data Center Builder for Network Architecture, Host Design

US-based Keysight Technologies has introduced the Keysight AI Data Center Builder, a software suite designed to validate and optimise AI infrastructure by emulating real-world workloads. This tool evaluates how new algorithms, components, and protocols impact AI training performance, the company announced. The KAI Data Center Builder integrates large language models (LLMs) and other AI model training workloads into the design and validation of AI infrastructure components, including networks, hosts, and accelerators. Keysight customers can access a library of LLM workloads like GPT and Llama, along with popular model partitioning schemas. Ram Periakaruppan, VP and general manager of network test and security solutions at Keysight, emphasised the importance of early validation in AI infrastructure design. "As AI infrastructure grows in scale and complexity, the need for full-stack validation and optimisation becomes crucial. To avoid costly delays and rework, it's essential to shift validation to earlier phases of the design and manufacturing cycle." The KAI Data Center Builder enhances AI training by aligning model partitioning strategies with AI cluster topology and configuration. Key considerations include the scale-up design of GPU interconnects, scale-out network design, and configuration of network load balancing. The solution reproduces network communication patterns of real-world AI training jobs to accelerate experimentation and provide insights into performance degradation. The KAI Data Center Builder is part of the KAI architecture, which offers end-to-end solutions for scaling AI data centers. Juniper Networks, a US-headquartered company in AI networking and cloud, has been one success story by improving its AI network efficiency with Keysight. "With Keysight, we are able to replicate the real data centre environment -- specifically for AI data centre requirements, such as RoCEv2," said Mahesh Subramaniam, director of product management at Juniper Networks.

Keysight Unveils Architecture for Scaling AI Data Centers

Enables AI infrastructure providers to emulate and optimize all aspects of the data center, from the physical layer through the application layer Validates and optimizes system-level performance, delivering seamless interoperability in AI data centers Identifies weak links that degrade AI data center performance Keysight Technologies, Inc. (NYSE: KEYS) introduces the Keysight Artificial Intelligence (KAI) architecture, a portfolio of end-to-end solutions designed to help customers scale artificial intelligence (AI) processing capacity in data centers by validating AI cluster components using real-world AI workload emulation. Providing system-level interoperability, performance, and efficiency insights, KAI helps operators maximize system performance and pinpoint performance issues not found when testing individual components. Scaling AI data centers requires testing throughout the design and build process -- every chip, cable, interconnect, switch, server, and graphics processing unit (GPU) needs to be validated at both the component and system level. Full-stack workload emulation complements physical layer testing, revealing insights not found when testing components alone. Customers can extract peak AI performance sooner, increasing capacity more quickly and maximizing the return on the billions spent on AI clusters. The KAI architecture enables AI providers, semiconductor fabricators, and network equipment manufacturers to: Accelerate design: Debug cutting-edge high-speed digital designs; meet or exceed the latest PCIe, DDR, and CXL standards. Accelerate development: Verify component-level compliance, including high-speed interconnects, cables, and chipsets, and validate workload performance at the system level. Accelerate deployment and operations: Validate and tune system-level performance across the entire data center, reducing the risk of workload failures by using end-to-end emulation to pinpoint system performance issues before deploying in production. The Keysight AI architecture, which includes the newly announced KAI Data Center Builder, features four portfolio suites that together address all aspects of AI data center design, from pre-silicon simulation through post-deployment system testing and troubleshooting. KAI Data Center Builder. Emulate high-scale AI workloads with measurable fidelity to improve system performance, predict and mitigate the impact of component failures, and optimize data center operations. KAI Compute. Optimize high-speed digital designs and pioneer next-generation AI chip development with a suite of AI-ready tools that includes electronic design automation, bit-error ratio testers, oscilloscopes, and arbitrary waveform generators. KAI Interconnect. Validate optical and electrical data paths to ensure scalable, high-speed connectivity up to 1.6T with a suite of AI-ready tools, including sampling oscilloscopes, photonic power meters, and network interconnect testers. KAI Network. Benchmark AI network performance, detect bottlenecks, and optimize AI workload distribution with a suite of AI-ready tools that includes AI workload emulators, distributed network traffic generators, and network traffic emulators. KAI Power. Optimize power efficiency and energy management across data center components with a suite of AI-ready tools that include oscilloscopes, power rail probes, and electronic design automation. Alan Weckel, Founder and Technology Analyst, 650 Group, said: "Accelerating design and deployment of next-generation AI/ML ASICs is key to unlocking the market as customers move from foundational training to agentic models. AI interconnect through scale-up, scale-out, and frontend networks will drive record 800GE and 1.6T port shipments over the next several years with one of the fastest innovation cycles to ever occur in the industry. The Keysight Artificial Intelligence architecture and KAI Data Center Builder solution will help customers scale and adapt to this new market opportunity." Ram Periakaruppan, Vice President and General Manager, Network Test & Security Solutions, Keysight, said: "Scaling AI data centers requires more than component-level validation. Interoperability, performance, and efficiency are system-wide metrics that can only be measured under real-world network conditions. Keysight's AI solutions integrate our deep experience in traffic emulation, component, and network compliance validation, and the latest industry standards to emulate every aspect of data center performance: compute, network, interconnect, and power to ensure AI infrastructure meets evolving demands."

Keysight Introduces AI Data Center Builder to Validate and Optimize Network Architecture and Host Design

Validates the performance of AI infrastructure by emulating real-world workloads Evaluates how new algorithms, components, and protocols improve the performance of AI training Adjusts and optimizes the parameters of both AI workloads and system infrastructure without investing in expensive large-scale deployments Keysight Technologies, Inc. (NYSE: KEYS) introduces Keysight AI (KAI) Data Center Builder, an advanced software suite that emulates real-world workloads to evaluate how new algorithms, components, and protocols impact the performance of AI training. KAI Data Center Builder's workload emulation capability integrates large language model (LLM) and other artificial intelligence (AI) model training workloads into the design and validation of AI infrastructure components - networks, hosts, and accelerators. This solution enables tighter synergy between hardware design, protocols, architectures, and AI training algorithms, boosting system performance. AI operators use various parallel processing strategies, also known as model partitioning, to accelerate AI model training. Aligning model partitioning with AI cluster topology and configuration enhances training performance. During the AI cluster design phase, critical questions are best answered through experimentation. Many of the questions focus on data movement efficiency between the graphics processing units (GPUs). Key considerations include: Scale-up design of GPU interconnects inside an AI host or rack Scale-out network design, including bandwidth per GPU and topology Configuration of network load balancing and congestion control Tuning of the training framework parameters The KAI Data Center Builder workload emulation solution reproduces network communication patterns of real-world AI training jobs to accelerate experimentation, reduce the learning curve necessary for proficiency, and provide deeper insights into the cause of performance degradation, which is challenging to achieve through real AI training jobs alone. Keysight customers can access a library of LLM workloads like GPT and Llama, with a selection of popular model partitioning schemas like Data Parallel (DP), Fully Sharded Data Parallel (FSDP), and three-dimensional (3D) parallelism. Using the workload emulation application in the KAI Data Center Builder enables AI operators to: Experiment with parallelism parameters, including partition sizes and their distribution over the available AI infrastructure (scheduling) Understand the impact of communications within and among partitions on overall job completion time (JCT) Identify low-performing collective operations and drill down to identify bottlenecks Analyze network utilization, tail latency, and congestion to understand the impact they have on JCT The KAI Data Center Builder's new workload emulation capabilities enable AI operators, GPU cloud providers, and infrastructure vendors to bring realistic AI workloads into their lab setups to validate the evolving designs of AI clusters and new components. They can also experiment to fine-tune model partitioning schemas, parameters, and algorithms to optimize the infrastructure and improve AI workload performance. Ram Periakaruppan, Vice President and General Manager, Network Test & Security Solutions, Keysight, said: "As AI infrastructure grows in scale and complexity, the need for full-stack validation and optimization becomes crucial. To avoid costly delays and rework, it's essential to shift validation to earlier phases of the design and manufacturing cycle. KAI Data Center Builder's workload emulation brings a new level of realism to AI component and system design, optimizing workloads for peak performance." KAI Data Center Builder is the foundation of the Keysight Artificial Intelligence (KAI) architecture, a portfolio of end-to-end solutions designed to help customers scale artificial intelligence processing capacity in data centers by validating AI cluster components using real-world AI workload emulation. Keysight will showcase KAI Data Center Builder and its workload emulation capabilities in booth #1301 at the OFC 2025 conference, April 1-3, at the Moscone Center, San Francisco, California.