KAYTUS has introduced a managed service aimed at keeping artificial intelligence data centres and hyperscale computing clusters running through hardware failures, as the breakneck expansion of AI infrastructure turns system uptime into a business-critical concern.
Announced from Singapore on September 28, the OEM AI Managed Service provides integrated hardware maintenance and technical support for AI data centres and large GPU clusters. According to the company, the service combines locally stocked critical replacement components, factory-grade diagnostic and repair capabilities, and certified field service engineers posted on site to complete complex hardware repairs in as little as four hours.
Why uptime is becoming the bottleneck
The launch comes as the industry confronts the operational reality of training and running AI at scale: hardware fails, and it fails often. Meta's technical report on its Llama 3.1 405B training run recorded 419 unexpected interruptions across 54 days on a cluster of 16,384 GPUs — roughly one every three hours — with about 78 per cent attributed to confirmed or suspected hardware failures.
KAYTUS says its onsite delivery model cuts hardware recovery time by more than 30 per cent relative to generally available industry service levels, with an average end-to-end incident resolution time of 12 hours per failed node. Faster fault isolation and hardware remediation, the company argues, preserves schedulable computing capacity and keeps workloads moving on clusters that can run into thousands of accelerators.
The economics behind the service are large. Gartner projects global AI spending will reach $2.67 trillion in 2026, including approximately $1.48 trillion allocated to AI infrastructure. As clusters scale, even brief outages translate into significant lost compute, making availability and recovery time key service-level indicators for operators.
The offering reflects a broader industry shift: after two years dominated by the race to build AI data centres, attention is moving to the less glamorous but increasingly urgent work of operating them reliably.
Sources: GlobeNewswire — KAYTUS Introduces AI Managed Service for Data Centers, The Manila Times — KAYTUS Introduces AI Managed Service for Data Centers
Announced from Singapore on September 28, the OEM AI Managed Service provides integrated hardware maintenance and technical support for AI data centres and large GPU clusters. According to the company, the service combines locally stocked critical replacement components, factory-grade diagnostic and repair capabilities, and certified field service engineers posted on site to complete complex hardware repairs in as little as four hours.
Why uptime is becoming the bottleneck
The launch comes as the industry confronts the operational reality of training and running AI at scale: hardware fails, and it fails often. Meta's technical report on its Llama 3.1 405B training run recorded 419 unexpected interruptions across 54 days on a cluster of 16,384 GPUs — roughly one every three hours — with about 78 per cent attributed to confirmed or suspected hardware failures.
KAYTUS says its onsite delivery model cuts hardware recovery time by more than 30 per cent relative to generally available industry service levels, with an average end-to-end incident resolution time of 12 hours per failed node. Faster fault isolation and hardware remediation, the company argues, preserves schedulable computing capacity and keeps workloads moving on clusters that can run into thousands of accelerators.
The economics behind the service are large. Gartner projects global AI spending will reach $2.67 trillion in 2026, including approximately $1.48 trillion allocated to AI infrastructure. As clusters scale, even brief outages translate into significant lost compute, making availability and recovery time key service-level indicators for operators.
The offering reflects a broader industry shift: after two years dominated by the race to build AI data centres, attention is moving to the less glamorous but increasingly urgent work of operating them reliably.
Sources: GlobeNewswire — KAYTUS Introduces AI Managed Service for Data Centers, The Manila Times — KAYTUS Introduces AI Managed Service for Data Centers
