Skip to Content

I assume most of you would have heard Bon Jovi song “It’s my life, it’s now or never”. Through most of yesterday’s meeting in Boston, this song was playing in my mind – SAP is definitely in the “I ain’t gonna be just a face in the crowd, you gonna hear my voice when  shout it out loud” frame of mind – in a good way:) . Considering the quantity and quality of tweets flying around, with a healthy dose of scepticism and sarcasm mixed in, I think ecosystem was also in a similar frame of mind. The result – a very fun day. Special thanks to Mentor herders Mark and Aslan for getting me and some other mentors there to participate – much appreciated.

 

It was a terribly long flight for me to BOS from PHX. But that hardly mattered on Tuesday night since the first order of business was dinner with my mentor buddies Dennis , Jon , Aslan and Michael . We had a great conversation about the “state of the union” regarding SAP, and sharpened our tools for the actual event.

 

Wednesday morning started with a breakfast meeting with SAP CIO, Oliver Bussman and his team. The conversation was mostly on mobility, and how CIO’s office works with CTO’s office. Oliver is one of the most articulate CIOs I have seen, and is a great ambassador for SAP. I think SAP will have a lot easier time with customers if they take Oliver along for sales calls and let him explain how things work internally at SAP. Aslan gave Oliver his mentor shirt, and we clicked a lot of photos.

 

I also got to meet two of my favorite people that I only knew virtually before – Ethan J and Dick Hirsch. Although we only got a little bit of time together, we had some great conversations on SAP technologies. And Michael Koch and I recorded a short video too, which I am waiting to see.

 

Next up was the keynote by Vishal Sikka, Executive Board member of SAP. Vishal is the most tech savvy guy I know of. And he wears his passion on his sleeve. Keynote was nothing to write home about – demos did not look convincing, content seemed more technical than most people could understand in the audience and so on. On the plus side, Vishal did talk about the many new HANA based applications that were released. But the theme seemed to be HANA is all about speed and cool technology. Medidata tried to present a business case which made sense, but generally the audience did not seem very impressed.

 

Right afterwards was a panel discussion – with representation from SAP, Medidata, IBM and Vishal. This would have been a GREAT panel for Teched, and I saw people sitting on either side of me furiously searching on google for “VBAP table”. That gave me a big chuckle.

 

At this point, I was fairly sure it was a huge waste of time to get into a plane and come all this way to attend this event. But that changed in 5 minutes. We had a second meeting for just bloggers and analysts, and what an amazing session that was. Vishal removed his coat, and was in a black tutleneck – and tweets were flying around on comparisons with similarly dressed people leading other companies.  Vishal was a lot more relaxed, and the quality of the second round was many times more than the keynote from my perspective. This time around, the demos were spot on – with easy to understand business scenarios to go with it.  I was particularly impressed with the backward looking time travel possible with HANA based analytics. I am assuming that this is possible for all data sets in a columnar store, since change history is readily available within the data, unlike in relational storage in current SAP apps. it was awesome – and was widely appreciated. Smart meter analytics was another top class presentation.

 

Performance benchmarks of HANA was published. This was a joint exercise with IBM, and the results were mighty impressive. Check this excellent blog for details. Looking into SAP HANA Performance Test

 

If I am not mistaken HANA will become the database of choice from its 1.5 version. And eventually, HANA will become the DB for all SAP. Customers can still choose to keep ORACLE, DB2 etc if they want, but SAP will probably say everything works best with migration to HANA as DB. It is going to be hard for the big DB vendors, since they all have serious licensing revenue flows from SAP. IBM, Microsoft, ORACLE – watch out !

 

I will let experts on mobility explain the integration of mobility with high power analytics. I am sure Dick or John or Michael will write on mobile, river and other cool aspects discussed in the meetings. You should definitely check out this excellent video – by the JD-OD.com team where they discuss HANA and other related  topics with Dick and John.  http://www.zdnet.com/blog/howlett/saps-hana-roadmap-does-it-convince/2948?tag=mantle_skin;content

 

To me, just one point is not clear – bandwidth is the big limiting factor for mobile devices. if we liberate all the ERP screens and make them available to mobile, can it work with full features on the type of bandwiidth we see around the world? I doubt it, but I am sure some one wlill enlighten me on this topic.

 

Integrated planning is getting a big facelift with HANA – and planning apps in general should be the first to make a business case for HANA at customer sites. This led to the question “What about BPC?”. Vishal answered without missing a beat that BPC will get all the innovations from HANA too. That was heartening to hear.

 

At lunch with Vishal, Dennis and Jon recorded a lively debate between Vishal and me on the topic of Watson and HANA. Here is the video. That was a lot of fun – at least for me 🙂 http://www.zdnet.com/blog/howlett/ibm-watson-and-sap-hana-faceoff/2946?tag=mantle_skin;content

 

Post lunch, we had a good discussion with Aiaz Kazi and team on a variety of topics covered this far in the day. I came away impressed at how well SAP takes feedback, and how Aiaz explained the thought process behind some of the decisions.

 

The last meeting of the day for me was a 1:1 with M.Sethu, who is a senior leader in the HANA team. Sethu and I got introduced last year over email, but this was the first meeting in person. Sethu is one of those guys who can go VERY deep into the inner workings of the system, but the next second can uplevel the discussion to answer a business question.Since I lead IBM’s advanced analytics competency in SAP space, I was thrilled to ask him all the questions I had. And I got some excellent answers.

 

My primary question was on how these new application will be extended. I would think that most customers will need to change the datamodel, and some front end screens. The specific question was  – will SAP still support an application if customer changes it. SAP is trying to figure out the long term lifecycle management of the applications built on HANA, but at a minimum we know that these things can be extended by customers. We should have a more concrete answer by SAPPHIRE. I am also waiting to check out the SDKs coming out for building apps.

 

I probably missed half of the important stuff, but I am sure we will see blogs from several people who can articulate better than me. For me – this was a terrific day, and I will do it again. But please please make the next big annoucement in the West coast…pretty please.

To report this post you need to login first.

9 Comments

You must be Logged on to comment or reply to a post.

  1. Derek Loranca
    Sounds like it was a really good session in Boston.  I’m definitely looking forward to what SAP has in store for it US HQ crowd here in Philly next week!! 

    Thanks for taking the time to blog about it…much appreciated!

    (0) 
  2. Bala Prabahar
    Hi Vijay,

    I enjoyed watching your discussions with Vishal. This blog again is great one; however all I see is hype. I guess Dennis’s blog “SAP’s HANA roadmap: does it convince?” discusses that issue really well: SAPPHIRE…or bust.
    I’m a bit surprised that SAP HANA performance test was conducted on 4TB DB with the largest table containing 600m records(460billion to 600million records?).  IMO this sends wrong signal about HANA’s true capabilities. 600m records is not a lot. Even first test should’ve been conducted on a much larger dataset to demonstrate the difference between HANA and BWA.

    Thanks,
    Bala

    (0) 
    1. Stephen Johannes
      I think the difference is that HANA will have larger scope than BWA.  The sizing numbers are so huge, to it offers a glimmer of hope that the cost be significantly less for those of us with a lot less data in our systems.

      Take care,

      Stephen

      (0) 
      1. Bala Prabahar
        Hello Stephen,

        In TechEd, I heard HANA was able to handle 450(or 460b) billion real time POS records. 450/460 billion records is definitely huge; this number is at least 750 times more than 600m records table used in the test. In addition, BWA was able to handle 550m records in 2007.
        I agree with your opinion that HANA will have larger scope than BWA; however the test results don’t support that. May be next test and/or forthcoming white paper would provide more details.

        Thanks and have a great weekend,
        Bala

        (0) 
        1. Stephen Johannes
          By scope I mean that BWA which serves a specialized scenario and does the “job” is not the same as HANA which intended to have a larger functional scope.

          If HANA was only designed to do what BWA did, then I would agree it’s not as impressive.  However the long-term roadmap which suggests the possibility of HANA at the foundation of all applications is exciting.

          Record size is not why people are excited, its the concept of generally applying the usage beyond the specialized scenario that is excited.  Plus the hardware costs in 2007 do not equal the costs in 2011.  That’s why this is exciting, because the costs are approaching mass “enterprise affordibility”.

          Take care,

          Stephen

          (0) 
    2. Witalij Rudnicki
      I’ve been trying to make some sense from published HANA performance results with no luck. I expected that complete whitepaper from Wintercorp will be published and clarify the thing, but after seeing modifications introduced into Wintercorp audit letter, then link to audit letter removed from Wintercorp website, then link to it removed from Dina Bitton’s weblog, I do not think we will ever see this white paper.
      I absolutely understand that to get meaningful results comparison we need to make sure that we are comparing apples to apples. But even assuming some huge error margin, I am trying to understand what SAP tried to prove announcing 10,042 queries/hour for HANA benchmark on 600M recs table comparing to 854,649 nav.steps/hour in standard BW’s BI-D benchmark where multiple users run queries on data in 10 InfoCubes which contain 2,500,000,000 records?
      It is almost 2 orders of magnitude, so I assume I am missing something in this comparison, but so far did not see any meaningful explanation for interpretation of the data.
      (0) 
  3. Vijay Vijayasankar Post author
    I appreciate the comments.
    It was a lot of fun, and it was awesome to catch up with all my buddies from around the world. A mini-teched like experience.

    Especially heartening to see was the way in which the various influencer programs converged – SAP definitely is doing a good job in breaking the boundaries.

    (0) 

Leave a Reply