GPU-as-a-service represents a potential enterprise revenue opportunity for operators
02 September 2024 | Research
Article | PDF (2 pages) | Cloud and AI Infrastructure| Enterprise Services
Enterprises need access to infrastructure that can support their growing number of AI workloads. Rather than deploying on-premises AI infrastructure, many enterprises are consuming GPU-as-a-service (GPUaaS), also known as AI infrastructure-as-a-service (AI IaaS). Operators may be able to capitalise on the growing demand from enterprises for GPUaaS – indeed, a small handful of operators have already announced and/or launched GPUaaS offerings and others may benefit from developing similar offers.
Author

Joseph Attwood
AnalystRelated items
Company profile
Deutsche Telekom: enterprise AI services
Survey report
Large business fixed voice and online meetings usage: 1Q 2025 business survey
Article
Sylva Developer Day and KubeCon 2025: operators’ interest in the Sylva project is growing