The convention was anticipated to attract 25,000 reside attendees, plus as many as 300,000 extra on-line, making it among the many most heavily-trafficked occasions CommScope has participated on this 12 months.
A number of recurring themes emerged in the course of the five-day present in San Jose, California. Amongst these had been the quickly growing want for capability, density, energy and cooling options to satisfy the demand that expands with every new technology of AI know-how.
Power effectivity was additionally a key matter as AI knowledge heart server cupboards transfer in the direction of a staggering 600 kW attract response to escalating demand for contemporary AI functions. The surge in complicated workloads supporting superior reasoning fashions—and the rise of extra autonomous agentic AI—are driving an unprecedented want for extra computational energy. For AI knowledge facilities, this requirement additionally comes with an pressing want for improved power effectivity and extra agility in infrastructure design.
CommScope’s options empower agility and scalability
Our sales space demonstrations and presentation supplied attendees (together with key stakeholders, ecosystem companions, and world hyperscale, cloud and telco clients) with sensible steerage for a variety of priorities. They included bettering power effectivity and enabling speedy deployment, flexibility and redundancy for capability migration.
Exhibitors at CommScope’s GTC sales space
The current launch of the Propel XFrame™ answer was properly acquired and generated vital pleasure across the CommScope sales space. Prospects, companions and key decisionmakers had been typically impressed by the manageability, density and adaptability of the answer, particularly when paired with the revolutionary FiberGuide® overhead raceway system.
Cupboard layouts are based mostly on the NVIDIA reference structure fashions for every AI cluster. For programs designers, this affords worth in that it allows off-site meeting of widespread constructing blocks for scaling knowledge facilities, delivering on-site manufacturing efficiencies. Mixed with optimum designs present in CommScope’s Knowledge Heart Cabling Options for AI Networks information, on-site deployment may be quicker and extra environment friendly.
Getting extra AI efficiency for fewer watts
On the subject of power effectivity, a current announcement by NVIDIA relating to the potential of co-packaged optics (CPO) designs has ignited curiosity within the energy efficiencies that CPOs present in comparison with legacy transceiver mixtures on the face of electronics. Eradicating the transceivers at every finish of a cable and changing them with passive MPO fiber connectors on the entrance of the community tools in every channel—mixed with some inside overhead discount—can assist scale back energy use per port. Multiplied over the tons of of hundreds of ports in a typical AI knowledge heart, the mixed power financial savings can develop into fairly vital.
When paired with superior fiber-optic cabling resembling CommScope’s extremely low-loss Propel™ options, NVIDIA’s just-launched photonic change know-how affords unimaginable potential to offer next-generation efficiency and substantial reductions in energy consumption.
Ken Corridor hosts CommScope’s GTC presentation session
Our presence at NVIDIA GTC 2025 was a robust demonstration of the function that CommScope performs in enabling knowledge heart transformation and deployment of AI factories throughout the globe. To be taught extra about CommScope’s options for NVIDIA AI factories, try our featured sources right here, the place you may also discover our complete information Knowledge Heart Cabling Options for NVIDIA AI Networks.
© 2025 CommScope, LLC. All rights reserved. CommScope and the CommScope emblem are registered logos of CommScope and/or its associates within the U.S. and different international locations. For added trademark data see https://www.commscope.com/logos. NVIDIA is a trademark of NVIDIA Company. All product names, logos and registered logos are property of their respective homeowners.
The convention was anticipated to attract 25,000 reside attendees, plus as many as 300,000 extra on-line, making it among the many most heavily-trafficked occasions CommScope has participated on this 12 months.
A number of recurring themes emerged in the course of the five-day present in San Jose, California. Amongst these had been the quickly growing want for capability, density, energy and cooling options to satisfy the demand that expands with every new technology of AI know-how.
Power effectivity was additionally a key matter as AI knowledge heart server cupboards transfer in the direction of a staggering 600 kW attract response to escalating demand for contemporary AI functions. The surge in complicated workloads supporting superior reasoning fashions—and the rise of extra autonomous agentic AI—are driving an unprecedented want for extra computational energy. For AI knowledge facilities, this requirement additionally comes with an pressing want for improved power effectivity and extra agility in infrastructure design.
CommScope’s options empower agility and scalability
Our sales space demonstrations and presentation supplied attendees (together with key stakeholders, ecosystem companions, and world hyperscale, cloud and telco clients) with sensible steerage for a variety of priorities. They included bettering power effectivity and enabling speedy deployment, flexibility and redundancy for capability migration.
Exhibitors at CommScope’s GTC sales space
The current launch of the Propel XFrame™ answer was properly acquired and generated vital pleasure across the CommScope sales space. Prospects, companions and key decisionmakers had been typically impressed by the manageability, density and adaptability of the answer, particularly when paired with the revolutionary FiberGuide® overhead raceway system.
Cupboard layouts are based mostly on the NVIDIA reference structure fashions for every AI cluster. For programs designers, this affords worth in that it allows off-site meeting of widespread constructing blocks for scaling knowledge facilities, delivering on-site manufacturing efficiencies. Mixed with optimum designs present in CommScope’s Knowledge Heart Cabling Options for AI Networks information, on-site deployment may be quicker and extra environment friendly.
Getting extra AI efficiency for fewer watts
On the subject of power effectivity, a current announcement by NVIDIA relating to the potential of co-packaged optics (CPO) designs has ignited curiosity within the energy efficiencies that CPOs present in comparison with legacy transceiver mixtures on the face of electronics. Eradicating the transceivers at every finish of a cable and changing them with passive MPO fiber connectors on the entrance of the community tools in every channel—mixed with some inside overhead discount—can assist scale back energy use per port. Multiplied over the tons of of hundreds of ports in a typical AI knowledge heart, the mixed power financial savings can develop into fairly vital.
When paired with superior fiber-optic cabling resembling CommScope’s extremely low-loss Propel™ options, NVIDIA’s just-launched photonic change know-how affords unimaginable potential to offer next-generation efficiency and substantial reductions in energy consumption.
Ken Corridor hosts CommScope’s GTC presentation session
Our presence at NVIDIA GTC 2025 was a robust demonstration of the function that CommScope performs in enabling knowledge heart transformation and deployment of AI factories throughout the globe. To be taught extra about CommScope’s options for NVIDIA AI factories, try our featured sources right here, the place you may also discover our complete information Knowledge Heart Cabling Options for NVIDIA AI Networks.
© 2025 CommScope, LLC. All rights reserved. CommScope and the CommScope emblem are registered logos of CommScope and/or its associates within the U.S. and different international locations. For added trademark data see https://www.commscope.com/logos. NVIDIA is a trademark of NVIDIA Company. All product names, logos and registered logos are property of their respective homeowners.
