Public cloud providers are increasingly focusing their attention on services that require immense computing power, such as artificial intelligence, big data analytics and engineering use cases — a trend that’s expected to benefit major players in the market.
International Data Corp. projects the worldwide performance-intensive computing-as-a-service to grow to $103.1 billion in 2027 from $22.3 billion in 2021. IDC sees this market as a “fast-developing category of the public cloud services offerings” as enterprises try to run mathematically intensive computations.
For big public cloud companies, the opportunities are two-fold: they’re both offering special cloud technology capable for running these intensive services and offering applications feature AI and big data analytics.
It’s another way of replacing what many enterprises once did in-house with cloud services, bolstering their bottom lines.
“Just look at product design in cars: that world has moved from having mostly physical components like wind tunnel testing or clay forms to being digital now,” Leo Leung, vice president for Oracle’s cloud infrastructure group, said.
The top cloud service providers — the largest being Amazon, Microsoft, Google, Alibaba and IBM — are all poised to take advantage of the shift within many industries towards adoption of AI and high-performance computing services, Peter Rutten, IDC's global research lead on performance-intensive computing solutions, said.
“All of them are heavily invested in this,” Rutten said.
Adoption of AI more than doubled from 2017 to 2022, according to a report from McKinsey, from 20% of executives reporting adopting AI in at least one business area to 50% in 2022. With breakout stars like ChatGPT emerging in the conversational AI space, though, that adoption could grow in coming years.
Here come the robots
Increasingly companies are trying to convert jobs long done by humans to tasks solved by software as well, the McKinsey report noted.
Rohit Badlaney, general manager for IBM’s industry cloud platform, said his company’s clients are aggressively consuming more of their cloud capacity for these high-performance workloads, particularly financial and life sciences companies. He said IBM, which spent years touting its supercomputers, has spent years hoping to capture more of this market.
The challenge for companies like IBM will be delivering more year after year without interruption, he said.
“Your capacity planning has to be frictionless,” Badlaney said.
Other barriers could be the complexity in manage hybrid technologies and the lack of enterprise around high-performance computing in many companies, the IDC report warns. It also warned of disrupted IT spending pans due to supply chain issues or an economic slowdown.
The promise cloud service providers are offering is scalability, Rutten said, getting these services working quickly when they’re needed. Purchasing in-house data centers can take months or years to deploy, he said, while cloud computing services can scale up or down as needed.
“Businesses prefer to run these workloads in the clouds because you spin them up when you need and take them off when you don’t,” Rutten said.