Newshot

Stalled data center buildouts send inference chips hunting for space in telco and colo sites

Summary

Demand for inference is outrunning the construction of new AI data centers, as power, capital and permitting delays push fresh capacity further out. That gap is sending chipmakers to hunt for room in existing facilities. Many older telco and colocation sites have the space and power. But most can’t take today’s dense, liquid-cooled GPU racks […] The post Stalled data center buildouts send inference chips hunting for space in telco and colo sites appeared first on SiliconANGLE .

Original text(via SiliconANGLE)

UPDATED 07:13 EST / NOVEMBER 18 2026

Stalled data center buildouts send inference chips hunting for space in telco and colo sites

by Jonathan Anthony

Demand for inference is outrunning the construction of new AI data centers, as power, capital and permitting delays push fresh capacity further out. That gap is sending chipmakers to hunt for room in existing facilities.

Many older telco and colocation sites have the space and power. But most can’t take today’s dense, liquid-cooled GPU racks without costly retrofits, which opens the door for air-cooled inference systems, according to Anton McGonnell (pictured), vice president of core product at SambaNova Systems Inc.

“We as an industry are just encumbered by the velocity of data center buildouts, and that’s been slowed down by just the physics of how long this takes, by the access to capital, and more and more by legislative hurdles — particularly in the U.S.,” McGonnell said.

McGonnell spoke with theCUBE Research’s John Furrier and Gemma Allen at SC26, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed AI data centers and the demand for air-cooled inference racks. (* Disclosure below.)

Reviving brownfield capacity for AI data centers

SambaNova’s fifth-generation chip, the SN50, targets the decode phase of inference, where output tokens are generated and data movement becomes the bottleneck. Its racks stay air-cooled and light enough to fit existing facilities, McGonnell explained. Newer GPU racks, by contrast, can demand reinforced flooring and 120 to 250 kilowatts of power per rack, he noted.

“On all of this stuff — the power and the cooling gear — there’s 18- to 24-month lead times to get this equipment into the data centers. Never mind all of the construction costs,” McGonnell said.

SambaNova, which investors valued at $11 billion in July, pairs its decode chips with partners that handle prefill, the compute-heavy first phase of inference. At SC26, it is demonstrating Intel’s D-Series GPUs for prefill alongside its SambaRack SN50 for decode in an air-cooled heterogeneous setup.

“They have developed a chip that has much, much better cost performance for the prefill phase of inference,” McGonnell said. “They’ve sacrificed HBM, which isn’t needed for prefill because it’s a compute-bound problem.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of SC26:

(* Disclosure: NYSE Wired sponsored this segment of theCUBE. Neither NYSE Wired nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

A message from John Furrier, co-founder of SiliconANGLE:

Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.

  • 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more
  • 11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network

Are you an AWS customer? Support SiliconANGLE financially by buying your AWS services from our Marketplace portal page and links: https://siliconangle.com/aws-marketplace/

Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.

  • Humyn Labs raises $47M to bridge the gap between AI training and real-world robotics
  • From store outlets to methane leaks, every scrap of power is now fair game for AI compute
  • Exclusive: Wherobots AI agents automate complex spatial data engineering
  • Scripts can't run AI factories: Dell moves cluster management from the server to the rack
  • CloudBackend grabs €6.6M to turn AI-generated prototypes into productive business applications
  • At Ignite 2026, Microsoft reinvents Copilot as the autonomous engine of enterprise work
  • Humyn Labs raises $47M to bridge the gap between AI training and real-world roboticsAI - BY KYT DOTSON . 13 MINUTES AGO

Humyn Labs raises $47M to bridge the gap between AI training and real-world robotics

AI - BY KYT DOTSON . 13 MINUTES AGO

  • From store outlets to methane leaks, every scrap of power is now fair game for AI computeINFRA - BY JONATHAN ANTHONY . 1 HOUR AGO

From store outlets to methane leaks, every scrap of power is now fair game for AI compute

INFRA - BY JONATHAN ANTHONY . 1 HOUR AGO

  • Exclusive: Wherobots AI agents automate complex spatial data engineeringBIG DATA - BY PAUL GILLIN . 2 HOURS AGO

Exclusive: Wherobots AI agents automate complex spatial data engineering

BIG DATA - BY PAUL GILLIN . 2 HOURS AGO

  • Scripts can't run AI factories: Dell moves cluster management from the server to the rackINFRA - BY JONATHAN ANTHONY . 2 HOURS AGO

Scripts can't run AI factories: Dell moves cluster management from the server to the rack

INFRA - BY JONATHAN ANTHONY . 2 HOURS AGO

  • CloudBackend grabs €6.6M to turn AI-generated prototypes into productive business applicationsAPPS - BY MIKE WHEATLEY . 7 HOURS AGO

CloudBackend grabs €6.6M to turn AI-generated prototypes into productive business applications

APPS - BY MIKE WHEATLEY . 7 HOURS AGO

  • At Ignite 2026, Microsoft reinvents Copilot as the autonomous engine of enterprise workAI - BY MIKE WHEATLEY . 8 HOURS AGO

At Ignite 2026, Microsoft reinvents Copilot as the autonomous engine of enterprise work

AI - BY MIKE WHEATLEY . 8 HOURS AGO

The text above was extracted automatically from the source page and may miss images, tables or code blocks. The original is authoritative.

Published by SiliconANGLE
This site only summarises and aggregates; copyright stays with the original author. Headlines and summaries follow the source — where they differ, the original wins.
Read the original ↗