The numbers don’t lie. There are big savings in time and money for biotechs and pharmaceutical companies that turn to cloud computing, but it may only be for a while, not necessarily for the long term.

Definitions vary for cloud computing, with each industry expert and service provider offering their own interpretation. The broadest view includes most hosted services — including storage, software as a service, grid computing and hosted applications — while a narrower definition says that the cloud is about on-demand access to compute and storage services. Across that spectrum, companies from startups to multinationals, are reaching out to service providers to get access to computing power and applications without the investment in money and staff to build out a computing and network infrastructure.

“Cloud computing is a big, trendy word right now, compared with other technologies out there because it is so new. There is a lot of potential for biotechs and pharmas, especially the up and coming ones who don’t have a lot of resources,” said Ruchi Mallya, pharmaceutical technology analyst with market research firm Ovum, a part of the Datamonitor Group.

Allan Louie, an analyst with IDC Health Insights in Framingham, said the IT startup costs aren’t just about hardware. “You’re not just buying a rack of servers to install in an IT facility. It’s the long-term management of that hardware, the personnel costs. If you are simply renting that capability, you don’t have to worry about that cost. You eliminate a large portion of your hardware budget, and you don’t have to hire people to manage it,” he said.

Startups would rather spend their limited funds on research than on infrastructure, and cloud computing opens up new possibilities for collaborations with partners such as universities, big pharma and disease foundations. While startups still need some IT infrastructure, moving some processing out to the cloud in an on-demand manner can save up to 50 percent in IT costs, said Louie.

So, analysts see some startup biotechs turning to services, rather than building out an IT infrastructure. In addition, many pharmaceutical companies have relied on cloud computing. However, in each case, there are limits to how long companies use the cloud. The cloud is fine for research and discovery, and to get a company off the ground. But when patient data comes into play, and the U.S. Food and Drug Administration gets involved, biotechs and pharma are turning inward to more traditional infrastructures.

“As you look at early stage research, you don’t have to worry about the regulatory requirements involved in later stage and clinical work,” said Louie, noting that in-house systems and the policies and procedures that govern them also are important when biotechs reach the stage where they have to think about mergers and acquisitions.

¬†Where cloud computing works best for pharmaceuticals is at the early research stage for ad hoc projects. “For big pharma, a lot of it is about cash and resources. If I use the Amazon platform for early discovery, it’s a lot more cost effective to use 100 of their computers to run simulations for four months than to buy them and manage them myself,” said Mallya.

Looking ahead, Mallya expects the FDA to offer companies more guidance on how to work with outside services, particularly where patients’ clinical data is concerned. It’s still early, she noted.