By Arvind Krishna
Chemists at Unilever, the Anglo-Dutch consumer products giant, used to spend up to three months in their laboratories creating new formulations for liquid cleaning products. Now, they can perform the same work in 45 minutes or less–thanks to a collaboration between Unilever, one of the United Kingdom’s national laboratories and IBM.
Unilever product developers use iPads to set up tests and experiments, run simulations on an IBM Blue Gene/Q supercomputer at the UK’s Hartree Centre lab, and see their results in 3D visualizations that help them explore the data and make discoveries that otherwise might elude them.
This is an example of what’s possible when government, businesses and tech companies combine forces to bring the power of supercomputing and sophisticated data analytics to bear on business problems. It’s also an example of the kind of collaboration I expect to see flourish as a result of an agreement IBM is announcing today with Britain’s Science & Technology Facility Council.
The British government is investing to bolster its drive to create an information economy and to use Big Data to help generate economic growth. For IBM’s part, we’ll design and install a new high-performance computer at the Hartree Centre lab; provide a wide variety of software, including Watson cognitive technologies delivered via a cloud computing service; and assign IBM Research scientists to collaborate directly with UK government and industry scientists on their research projects. In addition, IBM will provide consultants with deep industry and technology expertise to help the lab plan and execute projects.
We’re not just selling the UK a collection of computer hardware and software; we’re forming an innovation partnership aimed at helping the country harness the power of Big Data.
As nations around the world embark on efforts to boost economic growth and improve their global competitiveness, their ability to tap Big Data will be an essential success factor. Extracting insights from data has tremendous potential for improving decision-making in business, science and society.
It’s essential for governments and industry to work together on this. With government support, companies large and small can get access to high-performance computers and the most sophisticated analytical tools. By working together, government and industry scientists can create hardware designs, software tools and algorithms that are fine-tuned to take on the Big Data challenges.
In our tie-up with the British government, we’re building one of the first computers in the world architected from the ground up based on the principles of data-centric computing.
Big Data is comprised of all of the unstructured data in the world–digital documents, Web sites, social media interactions, photos and videos, plus sensor data from the Internet of Things. In the Big Data era, it requires too much money and electrical energy to move all that data around within networks and computing systems, so we have to bring the data and the processing closer together. That will be done both by moving some of the processing to where the data is stored and by caching more of the data that’s being worked on closer to the processors. In a nutshell, that’s data-centric computing.
For the British government, we’re building systems based on OpenPOWER technologies. OpenPOWER is an alliance of more than 100 companies that are building computing systems and components rooted in IBM’s POWER processor and related system technologies. By opening our intellectual property to others, we enable companies to share resources, collaborate and accelerate innovation. We see OpenPOWER as a foundation upon which companies can design data-centric computers.
In this case, IBM is teaming with NVIDIA, a designer of powerful specialized processors called GPUs, and Mellanox, a maker of technologies for speeding data between the components in a computing system. We’ll also get feedback from UK research leaders and industry people to help us tweak the system. This kind of feedback is essential, not just in the design of this particular system but to help shape decisions about design of our data-centric systems going forward.
I’m tremendously excited about this collaboration. It’s a large, strategic engagement; it’s a true partnership; and it showcases IBM’s broad portfolio of technologies and expertise.
The aim for IBM Research is to play an essential role for business and for society, and this hits the bull’s-eye.