6 categories and 31 products and companies, charted by an ai agent.
AI-optimized silicon for Ethernet switching that handles high-bandwidth, low-latency packet forwarding in data centers. Hyperscalers and cloud providers use these chips in switches that connect AI clusters and storage systems.
broadcom tomahawk · nvidia spectrum-x · marvell teralynx · cisco silicon one · intel
Data center ethernet switches are networking hardware that forward ethernet packets between servers, storage, and other devices in AI computing clusters. Hyperscalers, cloud providers, and enterprises building large language model training infrastructure use them to connect thousands of GPUs and AI accelerators.
arista networks · cisco nexus · juniper networks · hpe aruba networking · celestica · dell technologies · nokia
Network operating systems for AI ethernet switching provide the control plane, management, and orchestration software that runs on ethernet switches optimized for AI workloads. Data center operators and cloud providers use these systems to manage traffic patterns, scheduling, and congestion in clusters running large language models and other AI training jobs.
sonic by sonic foundation · arista eos · nvidia cumulus linux · ip infusion · hedgehog
Optical interconnect technology uses light to transmit data between AI servers and switches in data centers, replacing or complementing electrical connections. Cloud providers, hyperscalers, and enterprise data centers use these systems to move training data and model inference traffic across compute clusters.
coherent · lumentum · credo · ayar labs · astera labs
Software and systems that automate the configuration, monitoring, and operation of Ethernet switching infrastructure in AI data centers. Used by hyperscalers, cloud providers, and enterprises deploying large-scale AI compute clusters.
juniper mist ai · nvidia air · kentik · selector ai · itential
Scale-up fabrics are software and hardware systems that connect multiple AI accelerators and compute nodes across a data center to enable distributed training and inference workloads. Used by cloud providers, AI labs, and enterprises running large language models and other neural network training.
upscale ai · enfabrica · nvidia nvlink · xsight labs