Keysight Technologies unveiled a new portfolio of scale-up validation solutions aimed at AI data center operators grappling with bandwidth, latency, and interoperability constraints as compute clusters grow denser. The company introduced tools supporting UALink 200G, PCIe 7.0, PCIe 6.0, PCIe 5.0, and CXL 3, targeting faster validation and deployment of high-performance AI infrastructure.
AI-driven workloads continue to push scale-up architectures that rely on ultra-high-speed interconnects for accelerator-to-accelerator and memory connectivity. As hyperscalers and enterprises expand GPU clusters and composable infrastructure, signal integrity, protocol compliance, and interoperability testing across multiple standards have become critical gating factors. Keysight positions its new offerings to reduce debug cycles and accelerate compliance testing across link, transaction, and coherent memory layers.
The solutions integrate Keysight’s BERT, oscilloscopes, protocol exercisers, and analyzers to validate stressed-signal performance, transmitter compliance, and coherent memory pooling across heterogeneous compute environments. Keysight will demonstrate the portfolio at DesignCon 2026 in Santa Clara.
• UALink 200G Receiver Conformance Test Software – Validates 200 Gb/s interfaces using the M8050A 120 Gbd BERT and Infiniium UXR real-time oscilloscope; automates stressed-signal calibration and measurement coverage for conformance and debug.
• PCIe 7.0 Transmitter Test – Enables BASE specification compliance testing at up to 128 GT/s to support next-generation AI accelerators and networking devices.
• PCIe 6.0 Protocol Compliance Test Suite – Provides link and transaction layer validation at up to 64 GT/s using exerciser and analyzer hardware with automated execution and pass/fail reporting.
• CXL 3 Protocol Exerciser and Analyzer – Validates coherent memory pooling across CPUs, GPUs, and accelerators to support disaggregated, memory-intensive AI workloads.
• PCIe 5.0 Protocol Link and Transaction Compliance – Certified by PCI-SIG for endpoint and retimer testing; focuses on standards-based interoperability and reduced deployment risk.
“As AI demand outpaces how quickly operators can bring power and fiber online, the real hurdle is scaling high bandwidth, low latency interconnects without sacrificing efficiency or interoperability. Keysight’s scale-up solutions help the industry validate performance, energy efficiency, and reliability across standards before deployment,” said Dr. Joachim Peerlings, Vice President and General Manager, Network and Datacenter Solutions, Keysight.
🌐 Analysis: Keysight’s announcement aligns with the rapid transition to scale-up AI fabrics built around 200G SerDes, 128 GT/s PCIe 7.0 signaling, and coherent memory pooling. As companies such as NVIDIA, AMD, and hyperscalers push rack-scale and cluster-scale AI architectures, validation complexity increases across physical, link, and protocol layers. Early compliance and interoperability testing can reduce costly redesign cycles in multi-vendor AI ecosystems.
Keysight’s focus on UALink 200G and CXL 3 also reflects broader industry momentum toward open accelerator interconnects and memory disaggregation. As PCIe 7.0 and next-generation CXL deployments move closer to production silicon, test and measurement vendors will play a central role in qualifying high-speed links that underpin AI factory scale-out and scale-up designs.
🌐 We’re tracking the latest developments in networking silicon. Follow our ongoing coverage at: https://convergedigest.com/category/semiconductors/
🌐 We’re launching the “Data Center Networking for AI” series on NextGenInfra.io and inviting companies building real solutions—silicon, optics, fabrics, switches, software, orchestration—to share their views on video and in our expert report. To get involved, send a note to [email protected] or [email protected]
