AI AND DATA CENTRES
and where they run. Smaller models suit defined tasks, and decisions between cloud, on-premises and edge will come down to data access, latency, control and available power. “This should also shift more channel value into the operational phase. As AI moves from experimentation into production, customers will increasingly need partners not just to deploy environments, but to operate, secure and continuously optimise them. That creates recurring opportunities around managed AI operations, governance, performance monitoring and cost optimisation.” Mark says that we can expect accelerated investment in AI-optimised infrastructure, continued growth in hybrid and multi-cloud environments, and increased focus on data sovereignty and cyber resilience. “The successful parties will be those that can combine performance, security, governance and sustainability into a single strategic approach,” he adds. Anton Shelepchuk, VP of worldwide sales at NAKIVO, says that rapid AI adoption will push corporate IT infrastructure past traditional limits, creating immense data management bottlenecks and intense architectural fragmentation. “As enterprises struggle to process, store, and analyse massive AI workloads, channel businesses must shift away from one-time hardware sales to become hyper-specialised strategic advisors who actively help businesses overcome these underlying data centre challenges,” he says. “The channel must pivot toward automated managed solutions over transactional software sales. Supported by flexible vendor licensing and specialised integration tools, partners will successfully guide enterprises through complex infrastructure landscapes to deliver true resilience, compliance, and sustainability.” n
about where new capital investment genuinely delivers value. “For me, that is the real opportunity for the channel in 2027,” he says. “Competitive advantage won't come from simply selling more compute. It will come from helping customers extract more value and resilience from every watt, every litre of water and every piece of infrastructure they already own, while intelligently planning what comes next.” Matt William, head of data centres at Zayo Europe, says power availability is becoming a defining factor in where and how data centre capacity can be developed. “For businesses looking to expand their AI capabilities, energy must be considered alongside connectivity, location and capacity,” he says. “Businesses should assess how effectively their infrastructure uses available power, including the efficiency of computing equipment, cooling systems and the wider data centre environment. Selecting facilities with strong energy management practices can help customers make better use of available capacity while managing costs and environmental impact.” Trends AI continues to develop at pace, which means the issues already being faced are likely to grow. Dennis says that he expects that in 2027 there will be much closer scrutiny of what AI costs to run. “Consistent performance, manageable running costs and evidence of business benefit will count for more than a good demo,” he says. “Infrastructure will increasingly be designed around everyday inference, and customers will need to understand how response times and costs change as usage grows, especially where AI agents make several calls to complete one task. Proper governance will be more critical than ever, which means that we’ll also see more selective choices about models
Competitive advantage won't come from simply selling more compute. “ ”
Contributors
Dennis Frank
hitachivantara.com
Matt William
zayoeurope.com
Anton Shelepchuk
nakivo.com
www.newsinthechannel.co.uk
47
Powered by FlippingBook