ISG Provider Lens™ Multi Public Cloud Services - SAP HANA Infrastructure Services - Brazil 2024

GenAI has accelerated cloud adoption while service providers rapidly adapt to market changes

Generative AI (GenAI) dominates every IT discussion today. However, IT is under pressure to reduce total spending, making it difficult to justify and approve budget increases to accommodate GenAI project costs. Vendors and consulting firms also do not provide clear visibility on operating costs after embedding GenAI into business applications because
estimating how many prompts and tokens users will input when using GenAI for trivial activities is difficult.

FinOps continues to be at the top of enterprises’ priorities. However, FinOps tools concentrate on measuring virtual machine (VM) usage and database access without specific tools to monitor and measure GenAI utilization and cost.

The public cloud continues to expand in every region. Cloud providers report doubledigit revenue growth after several quarters of diminishing growth ratios. Fundamental reasons for expansions are unclear, although most market  analysts consider GenAI to be one of the main drivers. This study could not confirm this affirmative. Most MSPs report
that clients demand pilot projects to measure effectiveness, but the number of GenAI projects is negligible. This study was able to identify approximately 150 enterprises in Brazil that are using GenAI in business functions (not including chatbots). It is a significant number to illustrate that GenAI delivers real value, but it is still small compared with the million enterprises that operate in the country.

Providers in the Consulting and Transformation Services quadrant report increasing demand for multicloud  architectures that use the best cloud for each workload. The market is still in the process of truly understanding multicloud. Some clients expect that multicloud will allow them to move their applications and data dynamically among clouds based on price variations and discounts, which is impractical. Moving from one cloud to another  requires application refactoring, network configuration and data transfers that consume time and resources. In the real world, an architecture review considers all cloud options to design the best distribution under scale, performance,
available connectivity, network latency, security tools and software licensing limitations.

The Managed Services quadrant assesses the top managed service providers (MSPs), including global providers that operate in Brazil. In general, these MSPs are working to understand GenAI’s value to AIOps and FinOps, but they have not recognized the expansion in service scope that GenAI will demand.

AIOps includes ML to identify incidents and solutions recorded by analysts providing data insights to automate resolution. MSPs review these insights from time to time. GenAI semantic search can read incident solution documentation to help analyze new incidents, but the productivity gain is irrelevant. Only a few MSPs are creating large libraries of configuration templates, cloud architecture designs, product configuration manuals and vast  knowledge documentation to enable GenAI indexation for more accurate search and recommendation. Shortly, MSPs will create a prompt with clients’ requirements, and GenAI will write migration plans and suggest architecture designs for human review. MSPs working in this direction expressed frustration that the large language models (LLMs) require extensive fine-tuning and training before they meet their expectations.

Most enterprises do not understand the expansion in service scope that GenAI will demand. Today, MSPs manage virtual machines, CPU and memory utilization, databases, availability, data lakes and data access. However, they do not manage AI analytics but monitor it as another application. Unlike analytics, GenAI is a service billed at user  access level and utilization (number of prompts and tokens). MSPs will be required to monitor GenAI usage, security, privacy and design architecture solutions to optimize GenAI utilization and costs, including comparing GenAI options across different clouds.

The FinOps Services and Cloud Optimization quadrant identified more providers than the previous year, showing a growing interest and market expansion. Cloud providers use distinct terminology and parameters to measure utilization and provide services under different pricing models. A typical client finds it difficult to understand and consolidate cloud billing. FinOps optimization is even more complex than billing and encompasses price, cost and
expenses, with usage monitoring at its core. Clients expect to reduce their cloud expenses, and MSPs alert that strict focus on cost cutting can lead to underperforming applications, causing unwanted business impacts.

FinOps services provide detailed insights into cloud spending, enabling clients to identify and eliminate waste and optimize resource utilization. IT waste drives energy waste and consequent unnecessary carbon emissions. In other words, IT overspending contributes to global warming, and clients needing to comply with carbon reduction targets must measure IT consumption and waste. Some FinOps providers can measure and report carbon emissions monthly, providing auditable evidence of environmental, social and governance (ESG) compliance.

The Hyperscale Infrastructure and Platform Services quadrant shows minor changes from the previous year when GenAI was already dominating the public cloud competition. Cloud providers announced many new GenAI tools and updates, improving GenAI response accuracy and regional availability.

In September 2024, Oracle announced its partnership with AWS, replicating the terms of its previous relationship with Microsoft Azure and Google Cloud to offer Oracle products to all three hyperscalers (AWS, Microsoft Azure and Google Cloud). This configuration facilitates multicloud design. SAP also has made its products available on the three  platforms, seamlessly enabling multicloud adoption.

The public sector is increasingly interested in using cloud services as providers can adhere to data sovereignty requirements. All the leading four hyperscalers — AWS, Microsoft Azure, Google Cloud and Oracle Cloud — offer onpremises infrastructure capable of delivering full cloud services, ensuring strict access control, data residency and client-side encryption.

The leading four hyperscalers have increased their partner network in Brazil, attracting many independent software providers (ISVs) to develop solutions and expand cloud services nationwide. The rapid cloud expansion allows niche ISVs to accelerate growth and increase margins dramatically. Thus, more and more ISVs are interested in building cloud competencies.

To drive further cloud demand, the leading four hyperscalers are investing in training and education, in partnership with clients and MSPs, to educate users in AI, analytics and GenAI. According to the hyperscalers, GenAI and cloud adoption are currently limited by the number of experts available.

The SAP HANA Infrastructure Services quadrant shows that the market continues to evolve and expand at the clients’ pace, which is slower than SAP would prefer. No new providers entered the market this year. However, the increasing demand for data sovereignty and availability of on-premises cloud appliances offered by the leading four hyperscalers may change the game shortly. These appliances can host SAP-certified virtual machines to run SAP S/4HANA locally, and they have cloud extensions that fit the SAP licensing requirement for public cloud infrastructure. If hosting SAP S/4HANA on the edge becomes popular, other service providers might launch new offerings based on the white-label Oracle Cloud Infrastructure (Oracle Alloy).

Access to the full report requires a subscription to ISG Research. Please contact us for subscription inquiries.