SAP's HANA database: Big performance for big data

By  

Yesterday, Sybase's parent company, SAP, made a shocking claim that demolishes the conventional wisdom in the enterprise software industry. The claim: SAP is a database company. Not just any database company, but the database company for Big Data.

I know, I know. SAP is an enterprise resource planning (ERP) and business applications company. Its software runs on databases like those from Sybase and other companies. That's true. But the fact is, SAP is also database company.

Don't believe me? Well, in addition to having the products in its Sybase arsenal, SAP makes the fastest, most robust database available today: HANA. HANA's in-memory, massively parallel processing architecture makes it ideally suited for the era of Big Data. And I mean really Big Data.

In a recent independent benchmark HANA raced through a 100TB test database with 100 billion records. First, HANA achieved a 20x data compression level, which was remarkable. More impressive, though, was that with no caching, indexing, or materializing of the query results, the query responses were a mere 300 to 500 milliseconds. Compare this to some Oracle documentation that has claimed it was "lightning fast" at processing 100 million records in one second. HANA, then, can run 1,000 times more data in less than one-half the time than Oracle.

Beyond benchmarks, in the real world of Wall Street, one HANA application is using Sybase CEP (Complex Event Processing) to feed more than 2.1 million updates per second into the database. In a retail environment in Japan, one customer achieved 400,000 times performance improvement over its previous database environment. Adobe uses HANA to analyze customer data in real time and T-Mobile runs three HANA databases to analyze and reduce customer churn. It's stories like these that make HANA the fastest growing product in SAP history.

HANA portends to change not just perceptions about SAP as a database company, but real lives. Imagine, if you will, a cancer patient sitting in a doctor's office having a sample taken to analyze her genome. With those test results, medical professionals can prepare a targeted, personally designed treatment program for the patient. The problem is that today those test results take at least one month to complete. With HANA, the potential is that the doctor will get those results during the patient's initial office visit and can begin the treatment immediately.

Join us:
Facebook

Twitter

Pinterest

Tumblr

LinkedIn

Google+

Data CenterWhite Papers & Webcasts

Webcast On Demand

Cloud Knowledge Vault

Sponsor: HP and Intel®

See more White Papers | Webcasts

Answers - Powered by ITworld

Join us:
Facebook

Twitter

Pinterest

Tumblr

LinkedIn

Google+

Ask a Question
randomness