LAS VEGAS – Cary, N.C. based analytics software firm SAS Institute parted the curtain on SAS Viya, its new analytics and visualization architecture that will eventually serve as the basis for all of its solutions and help usher its customers into the cloud era.
As more enterprises move to a hybrid cloud model, supporting workloads both on private servers and on a public cloud service, SAS made clear that Viya could be deployed in either scenario. More than that, the Viya architecture will open up SAS’ analytics engine to a wider audience that will be able to call upon it to answer analytics questions on a pay-per-use basis, using open programming languages and APIs. Company executives announced the new architecture at the annual SAS Global Forum conference on Tuesday.
Jim Goodnight, the company’s founder and its president and CEO, announced Viya in his keynote, describing it as “parallel algorithms that live in the cloud.” Referencing its in-memory analytics capability, he said that the architecture supported fewer procedures, but was more powerful.
“Today if you want to build a logistics model, a neural network model and a predictive model, you have to load that data three times,” he said in an interview. “Now you can load that data once and it stays in-memory.”
SAS is kicking off its Viya architecture era with an early adopter program according to Mike Frost, senior product manager for cloud and platforms at SAS. The aim is to have a general release in August or shortly thereafter, with four applications offered on Viya: Visual Analytics, Visual Statistics, Visual Investigator, and Visual Data Mining and Machine Learning.
The early adopter program will be free, to help encourage companies to make the transition to the new architecture. SAS wants to see all of its customers move to Viya eventually, but at the same time isn’t rushing them off older releases, with plans for several more releases to its current single-processor based sofware charted out in the years ahead.
SAS sees Viya as a cloud-ready architecture that can be adopted by startups and established companies alike, Frost says. “It can help solve the business problem of making that leap to solvency and stability,” he says. “For the established companies, it can help keep pace with the competition that’s coming from every corner.”
As more enterprises take advantage of public cloud offerings, more mission critical data is going to be living off the company campus, says Ray Wang, principal analyst at Constellation Research. “So you’d better have a way to bring that data back,” he says.
SAS demonstrated that wouldn’t be an issue for users. An onstage demo at the Global Forum focused on how users could take advantage of SAS 9 and SAS Viya in an interoperable manner, passing data seamlessly between applications operating on the different architectures.
Viya also opens up SAS to popular programming languages including Python and Java, and supports new APIs including public REST APIs commonly used to deliver web and mobile services. This allows a per-usage model not previously available with SAS. In this scenario, a user could upload a data set and run analytics, predictive modeling, and forecasting on it for a fee, all without owning any SAS software.
“We’re going to lower the barrier to entry for those that want to get started,” Frost says.
Opening up the APIs will allow enterprises to build services that include a call into SAS’ analytics capabilities, Wang says. It’s a model that offers more flexiblity than the old database approach.
“I can reuse my SAS investment as an existing customer,” he says. “As a new customer, I have the capability to pay as you go.”
SAS says its early adopter program will be launched in May.