The Nimbus Pricing Index (NPI) has recorded the first rise in prices for cloud services since their debut, signaling a potential trend of price increases due to inflationary pressures and weaker revenue growth for cloud service providers (CSPs). While the impact is still marginal in most regions, some European countries have experienced significant price increases. Computing power, a major component of cloud billing, continues to be a focus for CSPs, who are looking for ways to boost revenue without direct price-list increases. However, pricing in other currencies, such as euros and pounds, has seen hikes due to foreign currency exchange rate changes. Additionally, there have been price increases for network egress and storage in specific locations. GenAI, a popular topic across industries, poses the question of whether to build or buy AI training workloads for cloud services. While a vendor solution is generally more affordable due to economies of scale and in-house GPU availability, the decision becomes less clear for low-volume models. As volume and complexity increase, building in-house becomes harder to justify in terms of cost. Vendor choice does not heavily impact overall cost, except for high-volume models. Companies are also considering switching from x86 to ARM-based computing architectures to reduce costs and improve energy efficiency. ARM services offer scalability and cost savings compared to x86, but there are limitations in terms of simultaneous actions and rearchitecting existing applications. The Boston Consulting Group (BCG) partners with businesses and society to tackle challenges and drive positive impact through a transformational approach. Overall, the text discusses the trend of price increases in cloud services, the impact of inflationary pressures and weaker revenue growth, the role of computing power in cloud billing, the effects of foreign currency exchange rate changes, the considerations for building or buying GenAI models, the cost implications of vendor solutions and the potential benefits of switching to ARM-based computing architectures.
| Signal | Change | 10y horizon | Driving force |
|---|---|---|---|
| Cloud service prices increase | Increase in cloud service prices | Cloud services become more expensive | Inflationary pressures and weaker revenue growth for cloud service providers (CSPs) |
| ARM-based cloud architecture | Adoption of ARM-based processors in cloud architecture | More companies switch to ARM-based processors for cloud services | Power efficiency, cost savings, scalability, support for usage spikes |
| GenAI deployment options | Comparing build vs buy options for AI deployment | More companies choose off-the-shelf AI models for cost savings | Significant economies of scale and in-house GPU availability for vendors |