Viavi Solutions has launched its new peripheral component interconnect express (PCIe) 6.0 protocol analysis testing platform at the Flash Memory Summit conference and exhibition held at the Santa Clara Convention Centre.
PCIe is a high-speed computer expansion bus standard that enables emerging applications in data centre/hyperscale, cloud, artificial intelligence/machine learning (AI/ML), quantum computing, IoT, automotive and defence industry markets. Operating at 64 gigatransfers per second (GT/s) data rates, PCIe 6.0 doubles the bandwidth and throughput of the previous PCIe 5.0 specification.
The new VIAVI PCIe 6.0 protocol analysis testing platform, part of the VIAVI Xgig family, will include analysers, exercisers and interposers. The company will also provide an option that allows its customers to capture 64 GT/s PCIe 6.0 traces on the existing Xgig PCIe 5.0 equipment to extend the life of their equipment investments during the early R&D phase.
A robust platform designed to provide deep insights, the new Xgig PCIe 6.0 platform will include familiar Xgig analysis capabilities including:
- Xgig software suite: Trace control, Expert, Serialytics and MLTT simplify data interpretation and performance analysis while enabling developers to more quickly identify, locate and resolve errors. Full-stack analysis for NVMe and CXL.
- CXL protocol support: Analyser/exerciser accommodates all CXL release specifications, including CXL 3.0 and future releases. Every aspect and property of exchanges can be searched, filtered and analysed using advanced Xgig software tools.
- Link Training and Status State Machine (LTSSM) override: Enables users to create unexpected states or control states with corner-case LTSSM testing that uncovers hidden problems, enabling rapid identification and troubleshooting.
- Auto-tuning: Optimised fine-tuning to shorten tuning time and minimise reboots, along with EQ memory to eliminate the need for re-tuning.
- Port bifurcation: Increases productivity of a single analysis chassis by enabling multiple analysis instances to be performed at the same time by local or remote users.