In his Third Law of Motion, Isaac Newton noted that “for every action, there is an equal and opposite reaction.” While the physics are quite different, you could say much the same today to characterize the relationship between the growing use of high performance computing and artificial intelligence and the growing use of cloud computing for HPC and AI workloads.
In simple terms, when organizations turn to HPC and AI to capitalize on huge datasets, IT leaders have to make choices about where to run those workloads. In years past, the default choice would have been on-premises infrastructure. But today, IT leaders are increasingly seeing the advantages of running HPC and AI workloads in public and private clouds.
That’s a key takeaway from a new Forrester Consulting thought leadership paper, “Hybrid Cloud: A Smart Choice For AI And HPC.” In this paper, commissioned by Dell Technologies and Intel, Forrester pointed to a clear shift toward the cloud for HPC and AI workloads.
This was one of the overarching findings from a survey of global IT decision makers in government, higher education, manufacturing, financial services, healthcare, and oil and gas organizations. Among survey respondents, 33 percent plan to run HPC services in the public cloud and 28 percent plan to use private cloud for the same in the coming year.
Let’s look at some of the other findings from this survey. The Forrester paper notes:
- Hybrid cloud is the new reality. “Firms are making increasingly more use of cloud — both public and private — especially when it comes to HPC and AI workloads. But this doesn’t mean all deployments are moving off-premises. In reality, respondents say their firms expect cloud use to increase moving forward, but most are still working with environments that combine both cloud and non-cloud infrastructure.”
- Firms must contend with security, latency, and application architecture issues while managing cost. “The majority of respondents report struggling with security issues, network delays, and necessary app rework that can derail projects and lead to increased costs. Though managing costs is a universal goal, firms need to put a heavier focus on cost optimization if they want to be successful.”
- Hybrid cloud can be the cure for what ails firms. “Cloud migration will always be difficult. However, respondents in our study find that deploying HPC and AI workloads specifically in a hybrid environment results in business benefits that directly solve for some of their biggest challenges, most notably: improved performance, lower costs, and better security.”
- Mature cloud use can transform your business. “Respondents say their firms are already reporting improvements to their bottom lines from their move to cloud. What’s even more encouraging? Some mature organizations have instituted formalized processes, compiled sufficient resources, and maintain consistent management to help them use cloud as a way to fundamentally transform their business. As a result, these firms enjoy even larger gains from their efforts.”
The bottom line here is pretty clear. As organizations look to find the IT infrastructure to run new HPC and AI workloads, they are increasingly looking to cloud environments. As Forrester notes, “We found that to optimize performance and ensure success, firms are deploying HPC and AI on cloud, and they have no plans to stop anytime soon.”
To learn more
For the full story, see the Forrester Consulting paper “Hybrid Cloud: A Smart Choice For AI And HPC.”