As seen with the latest quarterly earnings reports, the hyperscalers continue to post strong growth with their cloud platforms. Amazon’s AWS remains dominant, with sales increasing 39% to $16.1 billion. As for the No. 2 player, Microsoft, the software giant’s Azure business grew 50%. Then, there is Google, which saw its cloud operations generate $4.99 billion, up 45% on a year-over-year basis.
There is much room for growth for the hyperscalers. According to IDC (International Data Corp.), the spending on cloud technologies and services is forecasted to go from $706.6 billion in 2021 to $1.3 trillion by 2025.
Yet the market is getting more competitive, so the hyperscalers will need to seek out new corners of the cloud market. One is actually mainframes. With the urgency for digital transformation and the changes wrought from COVID-19—such as with the move towards more flexible work—large enterprises are more willing to rethink their legacy systems.
Let’s see what this means for the hyperscalers.
Why the Mainframe?
The mainframe market is enormous. Based on research from BMC, this technology is used by:
- All ten of the world’s largest insurers
- 92 of the world’s top 100 banks
- 18 of the top 25 retailers
- 70% of Fortune 500 companies
Mainframes often operate mission-critical operations for large enterprises, such as payroll, customer accounts, insurance claims, airline reservations, and credit card processing just to name a few.
IBM currently dominates the mainframe market. It not only sells the machines but also owns major software platforms. The main ones include the IMS and Db2 databases as well as CICS (Customer Information Control System), which manages sophisticated transaction processing.
“Hyperscalers are moving into the mainframe market as they recognize that the majority of leading businesses in finance, government, insurance, and communications continue to run mission-critical applications on the mainframe,” said Nicole Ritchie, head of product marketing for Software AG Mainframe Integration Solutions. “As often quoted from SHARE and within mainframe circles, 70% of enterprise data resides or originates on the mainframe and up to $3 trillion in daily commerce flows through mainframes.
“That’s a lot of business.”
And yes, the IT budgets are large. More importantly, senior managers are looking to make investments so as to be more agile and innovative to provide better customer experiences and fend off highly funded startups.
The Mainframe Efforts of the Hyperscalers
The hyperscalers are in a prime position to benefit from mainframe modernization. These companies have huge financial resources, powerful software, global cloud infrastructures, and thousands of talented engineers. There’s also the advantage of having many large customers that likely have mainframe installations.
All the hyperscalers have partnership programs for mainframe migrations. For example, there is the AWS Mainframe Migration Competency Program. This includes consulting and software partners like Advanced, Blue Age, Deloitte, Micro Focus, TSRI, Accenture, and Wipro.
Although, in the case of Google, the company acquired Cornerstone Technology in 2020, a provider of mainframe migration software. Google transformed this into the G4 platform, which can translate complex COBOL, PL/I and assembler programs into Java and microservices. This is then integrated with Kubernetes containers.
“There is an opportunity for hyperscalers for pull-through revenue,” said Rob Anderson, vice president of marketing and product for application modernization at Advanced. “Each offers myriad products that are cloud-native—locked away from being accessible by the mainframe.
“By moving the mainframe to the cloud, they get the consumption revenue of those workloads and a massive, complex environment to sell solutions into.”
Will the Mainframe Go Cloud?
There are a myriad of successful use cases for the partnership strategy. Just look at the example of the migration of the logistics system for the U.S. Air Force (USAF) to Microsoft Azure. The platform served over 260,000 users a day, and the U.S. General Accountability Office (GAO) indicated that it was one of the federal government’s top 10 most critical legacy systems in need of modernization.
The requirements were definitely difficult. There could not be extensive retraining for the users. There was also a need for high levels of security, availability, and recovery.
“The USAF project entailed re-platforming their Unisys mainframe system to the cloud, converting legacy code to COBOL.NET and migrating the data to Azure SQL,” said Scott Silk, CEO of Astadia (the company that participated in the project). “As a result, the Air Force was able to preserve their existing investments in the system while modernizing and preparing the way for further digital transformation.”
By being in the cloud, the USAF system benefited from having access to modern technologies, including analytics and AI. There were also lower operating costs.
“Historically, we all think of mainframe modernization projects as being complex, high-stakes undertakings,” said Silk. “With automation, we’re seeing that change dramatically.”
Yet automation is not always a simple solution. Cultural issues can easily derail complex projects, especially when there continues to be the use of old approaches. There is also the temptation to “boil the ocean.” For the most part, there needs to be realistic goals and the use of modern DevOps—which are not necessarily the case with mainframe projects.
Total migration is not necessarily the right path either.
“The reality is that many businesses are reluctant to give up the processing speeds, reliability and security of the IBM Z,” said Ritchie. “Hyperscalers would be well served to consider a hybrid approach where mainframe and cloud work together as a much more desirable state.”