Google: TensorFlow, Open Hardware and More on Collaboration
Beginner's guide for TensorFlow: The basics of Google's machine-learning library
It is an open-source, accelerated-math library designed to help developers build and train machine-learning models using a wide range of hardware — CPUs, GPUs, and even specialized chips such as TPUs (Tensor Processing Units).
Google extends lowRISC FOSSi partnership
Unlike proprietary processors, the design and instruction set architecture (ISA) for which are kept behind a typically expensive licence wall, free and open source silicon (FOSSi) does what it says on the tin: Projects like RISC-V provide both the ISA and key implementations under permissive licences, allowing anyone to use, modify, distribute, and commercialise the technology without a single license or royalty payment - including, in many cases, the ability to create a proprietary implementation, should they so choose.
Following on from the news that it was a founding member of the Linux Foundation's CHIPS Alliance, an industry group set up to 'host and curate high-quality open source code relevant to the design of silicon devices', Google has now announced that it is extending its existing partnership with the lowRISC project to include additional funding, support, and the appointment of two Google staffers as board members on the project.
Google wants an open source silicon community for chip design
As evidenced by Android and Chromium, Google has long been committed to open source software. The company now wants to foster a similar community for hardware and chip design, particularly open source silicon.
To Create Prosperity, Free Market Competition Isn’t Enough—You Need Collaboration Too
What’s ironic is that all of this communal activity isn’t driven by beret-wearing revolutionaries plotting in coffee houses, but by many of today’s most powerful and profit-driven corporations, who act not out of altruism, but self-interest. The fact is that technology firms today who do not actively participate in open source communities are at a significant competitive disadvantage.
For example, Chris DiBona, Director of Open Source at Google, once told me, “We released Android as an open source product because we knew that was the fastest way to grow adoption, which enabled us to preserve the relationships with customers for businesses like search, maps and Gmail.” That is the reality of today’s marketplace. You collaborate in order to compete effectively. Businesses that don’t accept that simple fact will find it difficult to survive.
Science’s commitment to communal effort is not at all new, but is a thread running deep in America’s long history of technological dominance. And it’s not all about private companies competing with each other, either: it’s about how the market can benefit from public investment. When Vannevar Bush submitted his famous report, “Science, The Endless Frontier,” to President Truman at the end of World War II, he argued that scientific discovery should be considered a public good crucial to the competitiveness of the nation. The crux of his argument is that such efforts build capacity through creating what he called “scientific capital” and pointed out that “New products and new processes do not appear full-grown. They are founded on new principles and new conceptions, which in turn are painstakingly developed by research in the purest realms of science.”