Every computer age has a default mode for businesses.
At the beginning of the cloud era, the rallying cry was to move everything to the cloud. Now, at the end of the cloud era and the beginning of the AI era, there are changes again. The “everything in the cloud” mantra is giving way to business reality in the age of AI.
Cloud costs and business needs clearly dictate a more flexible and resilient approach for businesses. This is particularly true for production AI. Companies should consider all options, including on-premises data centers (co-location with enterprise-owned equipment also counts here), as well as hybrid and cloud solutions. On-premise facilities can give the company greater control in the areas of security, data sovereignty, and costs.
All companies view AI as a driver for future business growth, agility and cost savings, including making decisions about where and how to locate their AI production. This is one of the first questions companies must answer to achieve their AI aspirations. The decision to go on-premises or to the cloud needs to be carefully thought out for each AI workload. But this decision is not a one-size-fits-all choice: companies must be flexible and allow for a hybrid approach.
Both business factors and IT factors influence the decision-making process. One of those key factors is the roadmap for AI. The AI roadmap, which spans five to seven years, is the north star for the decision-making process about where to run AI production workloads. It’s helpful to understand that most IT and business decisions have to do with risk management. Some of the risk considerations on the IT side are: risks related to security, data sovereignty, startup/operation costs, resilience/uptime, disaster recovery, and business continuity. For some of the business risks, it is about time to value, how AI can increase business efficiency, and the long-term costs of AI.
Historically, most IT professionals have not considered the idea of using an on-premises data center for AI workloads. The immediate assumption is that upgrading an existing aging data center or building a new one is simply cost prohibitive. But in reality, that is not always the case or the only option, as modern co-location centers can handle the load.
Additionally, there are places like France, where AI data centers are being built under the direction of the government that will serve as a place for small businesses to share AI infrastructure and for larger companies to use for co-location, where all companies in a given data center share the costs.
Furthermore, we witness that cloud computing, no matter how good it is, still involves risk aspects. Running on shared computers can be a data security issue. Additionally, customers have no idea who has physical or virtual access to the infrastructure, and cloud costs have never really gone down. Yes, there are more features available, but the price has only increased yet.
Even more importantly, the three major cloud providers, Google, Microsoft, and AWS, are based in the United States. For American companies, that’s not a big problem. But for international companies, the issues surrounding data sovereignty are quite real. Keeping data not only under direct corporate control but also covered by local data regulations is a strong motivation to explore alternatives. Geopolitical instability, climate change and the Covid-19 pandemic have shown that supply lines, data rules and the provision of necessary technology can be compromised very quickly.
A company that owns its own infrastructure can exercise complete control over who has access to that infrastructure, ensure that all local laws and regulations are followed without worrying about interference from a foreign court, and can assure its own customers that responsibility for customer data lies with them, not a third party.
Companies can also adjust the size of their AI infrastructure investment based on their AI roadmap and intended use. In times when money may be tight, companies can also stretch their investment in AI infrastructure by delaying upgrades and maintaining equipment and systems a little longer. Whereas in a cloud deployment…the bills never stop coming.
There are more factors that should be examined when evaluating AI workload deployments, including considerations of a production cloud AI installation, the merits of a hybrid installation, and other business and IT risk factors to consider.
Take a look at this sponsored whitepaper via the link below, where GlobalData takes a much more extensive and intensive look at why enterprises should adjust their thinking about where to deploy their AI production workloads.
“Navigating between risks and rewards: Where should enterprises run their AI workloads?” was originally created and published by Verdict, a brand owned by GlobalData.
The information contained on this site has been included in good faith for general information purposes only. It is not intended to be advice on which you should rely, and we make no representation or warranty, whether express or implied, as to its accuracy or completeness. You must obtain professional or specialist advice before taking, or refraining from, any action based on the content on our site.
