SAS Global Forum Link-O-Rama

Google Alerts firing all day like an Uzi.  Quick summary.

Cloudera hearts SAS.  Cloudera announced a “strategic alliance” with SAS today at SAS Global Forum.    According to the announcement,

Customers are now empowered to quickly and easily analyze their data in Hadoop by connecting SAS directly to their Cloudera-powered Big Data repositories.

Customers were also so empowered last year at this time, when SAS first released an access engine for Cloudera.  SAS/ACCESS for Hadoop enables a user to embed MapReduce, HiveQL and Pig statements in a SAS program.

Leveraging SAS data management tools with Hadoop’s open platform and parallel architecture, business analysts can instantly query data in Hadoop without additional training

This statement is true for business analysts who already know MapReduce, HiveQL and Pig.

Does this mean that SAS predictive analytics will someday run inside a Cloudera Hadoop distribution?  Don’t hold your breath on that.   SAS seems to be putting all of its R&D eggs in the in-memory basket.

SAS hearts Cloud.  After poo-pooing public cloud for years, SAS finally admits that cloud has some potential; hence the hoopla about SAS 9.4 being “cloud-ready”.   Note to SAS: all software is “cloud-ready” unless you deliberately build in obstacles, like a cumbersome license key, terms and conditions or pricing that makes it not worth doing.

Your software renewal fees at work.   The Umstead, captive “hotel, restaurant and spa” where SAS wines and dines customers and employees of the month, has lovely glass sculptures in custom pots personally selected by Mrs Goodnight, with the able assistance of the SAS art and scenic crew.

I ate at the Umstead once.  When it first opened, word came down that SAS employees weren’t supposed to stay there, because  it was way too fancy.  Then it seems they had a hard time filling the place, because there’s no good reason to hang out at the corner of Harrison and I-40 unless you’re a SAS employee or your SAS rep gives you a free ticket to come on down, hang out and watch the Visual Analytics demo or whatever.

Food was pretty good, a little better than the Bonefish Grill or Ruth’s Chris up the street.  Service was mannered, as if the young people were still learning where to put the salad fork, which is not the sort of thing they teach you at N.C. State.  The cocktail waitress had a tramp stamp.

SAS reboots High Performance Analytics Server.  The global user group for SAS High Performance Analytics Server can meet at one of the small tables in Starbucks at the Moscone Center.   Announced two years ago and launched seventeen months ago, as of this writing SAS still has no public success stories for the product, possibly because customers are unwilling to shell out a couple million in first year fees plus a couple more million for the appliance it runs in for a big sandbox.

As the press release puts it:

Each of the new products is laser-focused on analytic technique, including data mining, text mining, optimization, forecasting, statistics and econometrics, and useful across any industry.

Got that?  It’s “laser-focused”.  Sounds like SAS is repackaging the HPA stuff into smaller bundles, presumably with a lower price, which seems like a smart move.

SAS also plans to add the HPA algorithms into SAS/Stat, Analytics Pro and Enterprise Miner for deployment in legacy environments.  This is great news for people with tiny data sets who like to play with SAS.

In other HPA news, SAS announced support for Oracle Exadata, a move that will cause long faces at IBM, who were really, really optimistic recently that SAS would soon support HPA on IBM boxes.  Note to IBM: if you want SAS to run on your boxes, you have to buy the Gold Sponsorship.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.