Altair Delivers More Options for Running SAS Code
August 1, 2022 Alex Woodie
Companies that have legacy SAS code for analytic applications running on IBM i servers and other platforms but don’t want to pay the SAS Institute for the runtime have another option from Altair, which recently acquired World Programming.
SAS was the dominant language for analytics for decades, starting in the 1970s, when Jim Goodnight and his colleagues developed the language at North Carolina State University. Goodnight went on to found the SAS Institute, which provided runtime and support for the open SAS code and, in the process, became the dominant provider of SAS solutions.
The SAS hegemony has only started to crack in the big data age of the past 10 years. Today, Python is the dominant platform for writing analytic routines, and R also is popular, particularly in analytic realms. However, when it comes to private industry, there is a vast installed base of existing SAS code that is running workloads day to day on everything from Windows and Linux servers to IBM mainframes, AIX boxes, and even IBM i servers.
Much like the legacy RPG and COBOL code that tens of thousands of organizations eventually want to modernize, there are millions of lines of SAS code that resists easy solutions. Rewriting the existing SAS code in a more modern language is one option, but there are obstacles to that route — not the least of which is the fact that SAS logic runs very efficiently and can’t be easily re-generated in another syntax.
The low efficiency of code converters also is a factor in preventing SAS customers from easy solutions. The SAS Institute wants to move these legacy customers to its modern Viya platform, but that has its own set of difficulties.
One company that came up with a solution is World Programming. The company, which was recently acquired by Altair, created a runtime called WPS Analytics that allows existing SAS code to run outside of the SAS environment. That can help existing SAS customers keep running their SAS code, but without paying SAS for the runtime, says Mark Do Couto, senior vice president of data analytics at Altair.
“They built a SAS language compiler and the ability to run the SAS engine within their engine without the need for a SAS license at all,” Do Couto said.
Altair supports the WPS Analytics engine running on Power, including IBM i and AIX, as well as the System Z mainframe, while minimizing cost on those platforms, DuCoto said.
“It opens up a lot of opportunity for our customers in general,” he said. “They have legacy hardware they haven’t touched, because SAS would charge them for more compute power. We’re giving them the ability to stay on mainframe or Power.”
Alternatively, Altair has a partnership with Dell on data analytics appliance, Do Couto said. The customers can work with Altair to move the SAS workloads to Dell, he said.
The WPS Analytics compiler covers about 92 percent of the SAS syntax, Do Couto said. However, the portions of SAS code that don’t compile tend to be more obscure pieces of the language that are rarely used. Altair will work with customers to identify these portions of the code and rewrite them if necessary, he said.
There was litigation between SAS and World Programming. However, that has all been cleared up, freeing Altair to pursue more SAS runtime opportunities in the US.