In a pair of announcements released yesterday, Google and Intel reaffirmed a partnership in four areas where the two have previously collaborated: the Kubernetes container orchestration framework, machine learning, internet of things development, and security.
In all four cases, the improvements involve software or services created by Google, in conjunction with hardware or contributions created by Intel. These aren’t new collaborations—instead, the companies are doubling down on areas where they have complementary efforts.
At the top of the list is Kubernetes, making fresh headlines with a newly launched commercial venture to provide support. Kubernetes, of course, originated at Google, and Intel has provided optimizations so that it can run more smoothly on Intel hardware—what Google calls “improved feature transparency” in its blog post.
This fits into a long-standing pattern set in which Intel contributes features to open source projects. In turn, the projects will increase their empowerment by—or dependency on—features in Intel hardware.
Take the Intel Distribution for Python, a version of the language aimed at math- and science-oriented developers. It makes use of the MKL, a set of Intel-authored high-performance math extensions to increase speed. Now consider Google and Intel are accelerating the TensorFlow machine learning library via—you guessed it—the MKL.
The two companies are also working to safeguard the connections between Intel IoT edge devices and the Google Cloud Platform (GCP), plus other “security integrations between Intel hardware and GCP infrastructure.”
One way to think of these types of partnerships and project-support efforts: They stem from Intel’s ongoing efforts to reinvent itself as the PC market continues to contract. Refocusing on the datacenter isn’t a total solution, as datacenter CPU sales don’t compensate directly for decreased consumer CPU sales.
Thus, Intel has been trying to broaden its presence by producing new kinds of support hardware and making its presence more desirable in cutting-edge software projects with growing audiences.
Another area where Intel and Google could reap mutual benefit is Intel’s work with hardware-accelerated container technology: the Clear Containers project and the experimental OpenStack-like CIAO project that runs atop it. There’s a lot of room for shared effort in such projects between both companies—although of the two, Intel needs help far more than Google does.