SAP HANA – The Cloud’s Silver Lining?
By Sam Sliman, President, Optimal Solutions Integration
Migration to the cloud is not a question of if but a complex riddle comprised of when, what, which order, how much and what flavor. Facing an inevitable cloud future, CIOs are redesigning their IT strategies in a way that solves this riddle -- mapping out a cloud migration plan that maximizes the payoff from existing technology while capitalizing on the cost and resource benefits of cloud-based services.
For SAP customers, many factors influence a cloud strategy, including growth plans, existing technology investments, security and reliability concerns, available SAP and non-SAP on-demand offerings, industry trends and best practices, and cost/competitive pressures -- to name just a few.
SAP’s on-demand product roadmap strategy is predicated on the pragmatic assumption that most upper-midsize and large organizations will adopt a hybrid model that includes a mixture of on-premise and on-demand solutions.
Tightly wedded to (guiding even) SAP’s on-demand product roadmap strategy is SAP HANA -- a relatively new platform that leverages SAP’s breakthrough in-memory computing technology for high-performance, real-time data analytics.
SAP HANA’s impact on SAP’s cloud and overall product roadmap strategy will be profound. As stated by SAP CTO and executive board member Vishal Sikka during his keynote address at SAPPHIRE 2011, "We believe the future of the cloud is in fact an in-memory cloud […] You will see us revolutionize the entire [SAP] product portfolio based on [SAP HANA]."
By understanding the proven capabilities and the very real but as yet untapped potential of SAP HANA, SAP customers can better visualize and formulate a plan for how their business might best embrace cloud services in an iterative, strategic and value-driven manner.
HANA may very well prove to be a silver lining for SAP customers as the cloud-computing paradigm shift plays out -- driving deeper cost savings from cloud services; increasing the cloud’s analytic and BI capabilities by orders of magnitude; wringing latency from cloud-based services; and infusing cloud services with unprecedented reach and flexibility to stave off commoditization.
And it is here that the two streams of on-demand and in-memory meet to create some interesting possibilities.
A snapshot of SAP on-demand solutions and SAP HANA deployment models
SAP’s current cloud offerings generally available include SAP Business ByDesign for small businesses (and subsidiaries of large enterprises); SAP BusinessObjects BI OnDemand and SAP BusinessObjects BI OnDemand for Salesforce.com for midsized and large organizations; and SAP Sourcing OnDemand, SAP Carbon Impact OnDemand and SAP Sales OnDemand for large enterprise customers. SAP StreamWork, appropriate for companies of all sizes, is SAP's on-demand collaboration solution.
A selection of SAP’s on-demand solutions currently in the pipeline includes:
- SAP Career OnDemand
- SAP Travel OnDemand
- SAP Sourcing OnDemand
- SAP Supplier Management OnDemand
- SAP Contract Lifecycle Management OnDemand
SAP currently has around 550 Business ByDesign customers and plans to have more than a thousand BBD customers by year’s end. As of February 2010, SAP had more than 260,000 subscribers to its on-demand BI offerings.
As proof of concept in support of HANA’s ability to transform cloud services, all of SAP’s on-demand products, including SAP Business ByDesign, run on the SAP HANA database.
SAP HANA was first demoed at SAPPHIRE 2010 and released for general availability in June 2011. Deployment models include the SAP HANA appliance for on-premise and the SAP HANA application cloud for on-demand (in private beta now).
SAP launched its first application powered by SAP HANA appliance software -- BusinessObjects Strategic Workforce Planning -- in March 2011 along with plans for a slew of new ‘powered by HANA’ applications as well as plans to revamp existing SAP solutions with the power of in-memory computing.
SAP currently has around 67 referenceable SAP HANA customers and a $571 million sales pipeline. The SAP HANA pipeline is growing by approximately $14 million a week, making it the fastest-growing sales pipeline in the history of SAP.
According to Andreas Wolf, an analyst with Warburg Research GmbH in Hamburg, SAP could have 100 HANA customers by the end of 2011, with each paying an estimated $1.5 installation fee.
Data growth and the server/database burden
Reducing capex is the foremost cloud driver. So long as reliability and security concerns are sufficiently addressed, offloading datacenters full of expensive, costly to maintain servers/databases to the public cloud (or reducing corporate server/database sprawl via private cloud) is a no brainer. Exponential data growth -- and the consequent need for more servers/databases to store and process this data -- adds significant momentum to this already compelling cost-savings driver. Big data means big hardware. Big hardware means big costs.
According to Gartner, enterprise data will increase 650 percent over the next five years. IDC pegs enterprise data doubling every 18 months, and MGI estimates that enterprises globally stored more than 7 exabytes - thousands of petabytes - of new data in 2010 alone. Given this explosive data growth, one can understand why cloud adoption is rapidly transitioning from an attractive money-saving option to an inevitable cost-saving imperative. It simply will cost too much for businesses to build, maintain, integrate and scale data centers capable of dealing with this explosive growth in data volumes.
Cloud migration, obviously, lightens the server/database burden for businesses, but what about the server/database burden for providers of cloud services (public and private)? At the end of the day, regardless of where they reside, an increasing number of servers/databases will be needed to sufficiently handle exponentially increasing data, and at some point, despite advances in server virtualization, data compression and MPP technologies, this will drive up cloud infrastructure costs, which in turn will drive up the cost of cloud services.
Following this logic, unless we see a tectonic shift in hardware technologies and database software, the CPU, memory and cost constraints of traditional datacenter infrastructure will eventually cap (and even diminish) the cloud’s value proposition of reduced costs.
Enter SAP HANA featuring SAP’s breakthrough in-memory technology. At bottom, SAP HANA serves as a database (albeit of a new order), and SAP is going directly after the database market with HANA, particularly traditional disk-based database management systems (DBMS).
By storing data in-memory rather than on hard disks, as traditional DBMS servers do, HANA servers eliminate the need for traditional, expensive, large-scale disk systems, cost less than 10 percent of these behemoth, CPU hogs, and perhaps most importantly, offer a viable solution to the big-data dilemma.
According to Rainer Zinow, SAP’s product strategy chief for Business ByDesign and the On Demand family of enterprise cloud applications, with HANA, “the future model of computing is based on infinite CPU cycles and infinite memory […] In a cloud computing world [utilizing HANA servers], CPU cycles don’t matter any more.”
Real-time analysis of big-time data for immediate adjustments and spot-on predictions
Storing the exploding volume of historical and real-time transactional data is one challenge. Cost-effectively extracting business value in real time from big data is quite another.
In this age of explosive data growth, businesses best able to convert raw, high-volume data -- structured/unstructured/operational/transactional -- into accurate, actionable information in real time will benefit from transformative operational insights, flexible opportunistic business strategy and sustainable competitive advantage.
The challenges here are speed and power -- or better stated a lack thereof. Cloud services based on traditional architectural approaches and handicapped by disk-based DBMS cannot process large data volumes fast enough to provide business users with the information they need to make real-time defensive and/or opportunistic adjustments to business processes. Unchecked, latency -- right up there with reliability and security -- is a cloud killer.
“The purity of [SAP’s] architectural approach with HANA gives us the ability to be released from the burdensome architectural requirements of the past, ” says SAP executive vice president Steve Lucas. “[HANA] was built specifically to be the platform that can not only handle but also wring maximum value out of Extreme Data.” (Extreme Data is SAP’s term for data volume even larger than big data.)
By storing data in-memory rather than on hard disks, HANA performs queries and analytics 100x – 1000x faster than traditional systems. This speed, coupled with HANA’s inherent ability to handle massive data volumes, dramatically broadens the possibilities for cloud-based services and applications.
According to Lucas, “SAP HANA is a platform for planning, forecasting, analyzing, storing, and eventually for OLTP at that extreme level.”
Loosely translated, this means that SAP HANA blows the lid off of today’s limited, ‘after-the-fact’, cloud services and unleashes limitless possibilities that enable intelligent interaction at the point of transaction and provide timely, accurate predictive capabilities for the next move a business should make.
For more on SAP HANA, read:
SAP HANA: Hype to Here in Less Than a Year
SAP’s Pragmatic Appliance Strategy
Analyze This – The SAP Way
In-Memory Technology and Enterprise Mobility
SAPPHIRE NOW's Most Memorable First
