Transforming database technology for faster big data insights – IBM and Intel

I started my career as a database administrator at Sequent computers, where I helped with the care and nurturing of the company’s Informix* database deployment. At the time, we considered it a real workhorse: It ran our corporate manufacturing database and ERP, plus supported our sales organization.

So it’s with many years of fond acquaintance that I note that Informix has just undergone a new release with version 12.1. Another foundational IBM database, IBM DB2*, has also just been re-released, in version 10.5. Additionally, IBM also upgraded IBM InfoSphere BigInsights* in their 2.1 version for Big Data. These new releases are more than incremental updates—each have been re-architected to massively accelerate analytic querying in particular, and information processing in general. This is particularly important news for IT organizations that face the challenge of spinning through mountains of data in search of hidden, business-changing insights and intelligence.

These new database releases from IBM build on close collaboration with Intel. Intel processors and software expertise, combined with IBM software co-engineered to make the most of Intel technology, have resulted in these truly revolutionary IBM database solutions.

Intel and IBM have worked together for decades to ensure that our customers get maximum performance and value from their investments. It’s never more urgent than now, when the challenges of big data require exceptional levels of processing power to extract intelligence from ever-growing mountains of raw data—arriving every day, faster, and from more diverse sources.

The new IBM DB2 release comes pre-integrated with BLU Acceleration, which adds the speed of columnar data processing to DB2’s traditional row/column-oriented data store. Columnar data processing is a much faster technology for scanning through massive data sets and performing analytical querying, while row store processing is faster for transactional computing. BLU Acceleration takes advantage of vector processing in multi-core Intel® processors to handle the column-based data, resulting in superfast analytics. What’s unique with the new release of DB2 is that the one database can perform both functions—in fact, you can draw on both row and columnar data at the same time to get answers that pull from both processing functions simultaneously, and at blinding speeds.

Columnar data is also much easier to compress, and the columnar format also enables unique in-memory technologies that stores frequently used data in RAM for instantaneous access, with less-frequently-accessed data stored on disk, where it remains compressed. It all adds up to speed-of-thought analytics and more cost-effective memory and storage systems.

In tests, query processing performance with DB2 10.5 with BLU Acceleration achieved 25x improvement over the previous release, DB2 10.1. To obtain these amazing gains in speed, IBM took advantage of the Advanced Vector Extensions (AVX/AVX2) instruction set on Intel® Xeon® processor E5-based systems. In addition, the new DB2 release uses the Single Instruction Multiple Data (SIMD) instruction set in Intel® Xeon® processors to pack multiple data elements into a single register. A single instruction can act on several data elements at once, munching right through masses of raw information in record time.

For more information on the science behind DB2 with BLU Acceleration, check out the article Borrowing from Science to Revolutionize Business Analytics by Chris Young.

The performance gains made on the new release of IBM Informix is even more astounding. Tests achieved a greater than 60 percent hardware performance improvement in query processing performance using Informix v12.1 when scaling from four socket Intel® Xeon® E7-4870 to eight socket Intel® Xeon® E7-8870 on the IBM System* x3850 server platform. In addition, the IBM Informix team has achieved a more than 500 percent software performance improvement using a geomean of 15 operational analytics from Informix v11.7 to Informix v12.1.

These new IBM database releases aren’t just about scaling performance. These platforms deliver higher power at a lower cost per query—they reduce costs by making the most efficient use of your system resources.

That’s the kind of performance gain you can take to the bank.

Follow Tim on Twitter @TimIntel.

Published on Categories Archive
Tim Allen

About Tim Allen

Tim is a strategic relationship manager for Intel driving enablement for enterprise software companies related to the cloud, big data, analytics, AEC, commercial VR, datacenter, and IoT. Tim has 20+ years of industry experience including work as a systems analyst, developer, system administrator, enterprise systems trainer, product marketing engineer, and marketing program manager. Prior to Intel Tim worked at IBM, Tektronix, Intersolv, Sequent and Con-Way Logistics. Tim holds a BSEE in computer engineering from BYU and an MBA in finance from the University of Portland. Specialties include - PMP, MCSE, CNA, HP-UX, AIX, Shell, Perl, C++