STAMFORD, Conn.—Spectrum plans to demo new capabilities to its Edge Compute Infrastructure (ECI) enhanced with NVIDIA accelerated computing platforms at the SCTE TechExpo, Sept. 29-Oct. 1 at the Georgia World Congress Center in Atlanta.
As AI moves beyond digital tasks to influencing and interacting with the physical world, the need for immediate, precise interactions with compute becomes critical. Spectrum says this increased need to use AI more quickly to control physical machines - robots, sensors, real-time video, secure personal data where geography adds too much latency has become “mission critical” and that it has begun to activate computing power across its network. This expansion is allowing its network to reach over 1,000 smaller facilities already embedded in its existing infrastructure—placing distributed compute capacity within 10 milliseconds of 500 million devices in homes and businesses across the U.S.
By combining its high-capacity fiber network and distributed footprint of powered edge data centers with NVIDIA accelerated computing, ECI brings AI infrastructure closer to where data is generated and decisions are made, the company said.
At SCTE TechExpo 26, Spectrum’s booth (B335) will host demonstrations stemming from close collaborations with Cast AI, HP and World Wide Technology (WWT), each illustrating how edge network capabilities are advancing cloud, AI and customer applications on Spectrum’s ECI platform. Spectrum also is working with Hydra Host to commercialize its AI Factory “Operating System” on ECI. These collaborations demonstrate Spectrum’s focus on building the distributed infrastructure that supports AI and data-driven technologies for a range of industries, the company said.
Cast AI will highlight its OMNI technology integrating NVIDIA Blackwell Ultra-based accelerated computing platforms, embedded, owned, hosted and operated by Spectrum on its ECI platform, into diverse hyperscale cloud environments. OMNI extends these GPUs natively into its multi-cloud compute fabric, where they appear as standard cluster nodes alongside major cloud provider capacity. Connectivity for this AI infrastructure is delivered through Spectrum's network, demonstrating how deep integration of compute and connectivity at the edge can advance AI innovation.
HP will showcase HP Z Boost, which shifts GPU access from hardware-bound to on-demand, accelerating rendering, visualization, and AI workloads through shared GPU resources. By delivering performance when and where it is needed, HP Z Boost helps organizations maximize GPU utilization while reducing infrastructure inefficiencies. HP will also demonstrate ZGX Nano and ZGX Fury, AI Stations purpose-built for enterprise edge AI development, enabling organizations to deploy and scale AI workloads closer to where data is created while reducing reliance on the cloud.
Hydra Host and Spectrum have an agreement allowing Hydra’s AI Factory platform to use capacity that is owned, hosted and operated by Spectrum on its ECI. As a global AI compute provider serving companies in need of GPU resources, Hydra Host will help extend the reach of Spectrum’s edge compute capabilities to organizations looking for advanced AI infrastructure solutions, Spectrum said.
WWT, NVIDIA and Spectrum are working together to advance distributed edge AI infrastructure for physical AI, combining Spectrum’s ECI with WWT’s system-integration expertise to help enable new applications such as robotics and industrial automation. WWT will exhibit a Unitree G1 humanoid robot in Spectrum’s booth, demonstrating advanced automation and kinetic AI as it performs dynamic tasks that showcase how edge compute enables scaled, low-latency, private and secure processing for next-generation robotics.
