SynopsisNvidia plans server price increases due to a global chip shortage. This shortage is expected to persist beyond the year 2027. Chipmakers are prioritizing AI-grade chips, impacting consumer electronics production. Rising server costs will increase AI model training and deployment expenses. Businesses face higher capital needs for extensive AI infrastructure development.ET BureauNvidia’s server price hike signals rising costsNvidia is reportedly planning to increase the price of servers containing AI chips - some customers being told of an increase of more than 15% - due to a global shortage of memory and storage chips that is likely to extend beyond 2027. The surge in prices of semiconductors is being fuelled by AI hyperscalers.Chipmakers Samsung, Micron and SK Hynix have pivoted to producing more AI-grade chips than those that go into consumer electronics. They are setting up new capacities. However, semiconductor fabrication plants take years to build, and new production is unlikely to come on stream until late 2028.Also read: Nvidia customers notified about AI-related price hikes above 15%: BloombergExport restrictions and supply chain diversion are adding to complexity by delaying the distribution of chip-making machinery.Nvidia will be joining a growing list of technology companies like Apple and Microsoft in failing to hold their product price lines because of the chronic shortage of chips. Advanced chips made by Nvidia are subject to US export controls, and the additional cost of its servers should have a bearing on how the AI race is being run.AI companies are running into energy and compute constraints as they intensify their infra buildup. They are also facing resistance from local communities over resource requirements of data centres. Rising costs of server farms will raise the capital intensity of training, deploying and scaling AI models. These costs will be transferred to client enterprises that are coming to grips with the sticker shock of AI.Some estimates say the chip shortage could extend until the end of the decade, with demand accelerating for AI solutions. By 2030, businesses will have shifted from basic AI bots to multi-agent systems embedded in global supply chains.About 1/3rd of current work hours could be automated over the next 5 yrs, requiring enormous investment in AI infrastructure. As the world's most valuable company - north of $5 tn - Nvidia holds a special place in this ecosystem, and its ability to absorb the rise in semiconductor prices will determine the pace of AI infrastructure growth. ...moreElevate your knowledge and leadership skills at a cost cheaper than your daily tea.Subscribe Now