On Aug. 6, Google and IBM announced a new open development alliance based around IBM’s POWER microprocessors. The two tech giants, along with graphics processor maker NVIDIA, network hardware company Mellanox and Taiwanese server manufacturer Tyan, are founding the OpenPOWER Consortium.
The consortium, according to Brad McCredie, CTO of the IBM’s systems and technology group, is looking to make hardware and software innovations for the future of massive scale and cloud data centers.
One of the first efforts, McCredie said, will be to integrate NVIDIA’s graphics processors with IBM’s microprocessors to squeeze out better performance and lower costs. Reducing the cost of operating a huge data center will only make the big data technologies currently being released more accessible, he said.
In this interview with Data Informed staff writer Ian B. Murphy, McCredie discusses the benefits and drawbacks of opening IBM’s POWER microprocessors as as a platform available to anyone, trying to engineer reliability into cloud data centers, and what he hopes this project will accomplish by next year.
Related articles on Data Informed: