Pentagon May Dump $10 Billion JEDI Program Over Microsoft, Amazon Fight

Pentagon May Dump $10 Billion JEDI Program Over Microsoft, Amazon Fight

The federal government is so tired of the lawsuit battle between Microsoft and Amazon over the $10 billion Joint Enterprise Defense Initiative (JEDI), it’s telling the court it may just kill the project rather than continue to litigate the matter.

The original JEDI program was supposed to provide the government with a cloud computing platform that would sit atop all other cloud deployments at the DoD and provide a general-purpose cloud environment. Smaller, specific clouds — what the government called “fit for purpose” clouds — were only to be utilized on an as-needed basis.

Pentagon May Dump $10 Billion JEDI Program Over Microsoft, Amazon Fight

The original JEDI contract was expected to go to Amazon until Oracle injected itself into the process, alleging improprieties. Former President Trump ordered a review of the project. Several months later, in October 2019, the government announced Microsoft, not Amazon, had won the project. Amazon promptly sued, alleging improper interference on the part of the government, based on President Trump’s longstanding and oft-expressed dislike for Jeff Bezos, the CEO of Amazon. Between Oracle and Amazon, the deployment of the JEDI system has been delayed for at least a year.

At this point, the government is ready to wash its hands of the entire affair. JEDI was supposed to be online over a year ago, and the government has had success deploying other cloud systems for smaller purposes in the interim, raising the question of whether JEDI is needed at all.

“If the court denies the government’s motion we will most likely be facing an even longer litigation process,” Pentagon spokesman John Kirby said. “The DOD Chief Information Officer will reassess the strategy going forward.”

The DoD has pivoted from insisting on the supreme importance of JEDI to talking up the other 13 cloud projects it has in play with various firms. While these were supposed to be organized under the JEDI general cloud, it may be that a different structure will address the same needs as effectively.

From the DoD’s perspective, dropping the project is probably the smartest move. JEDI can’t come online until the litigation is settled. That could take 2-4 years given how slowly the courts can move; Oracle and Google have been tangled in a lawsuit over whether APIs can be copyrighted for over a decade. The entire point of the JEDI project was to speed the deployment of cloud services across the DoD. Spending the next few years in court fighting for the right to deploy the program is exactly the opposite of how it was supposed to work in the first place.

Feature image by Touch of Light, Wikipedia, CC BY-SA 4.0

Continue reading

Google Told Stadia Developers They Were Making ‘Great Progress,’ Then Fired Them
Google Told Stadia Developers They Were Making ‘Great Progress,’ Then Fired Them

Google told its Stadia developers they were making "great progress." Then it fired them, less than a week later.

iFixit: Samsung ‘Ruined’ Its Smartphone Upcycling Program
iFixit: Samsung ‘Ruined’ Its Smartphone Upcycling Program

Samsung just dipped its toe in the waters of upcycling, but iFixit says Samsung lost its nerve, and that the original version of Galaxy Upcycling was going to be much more ambitious and useful.

Microsoft Kicks Unsupported PCs From Windows 11 Testing Program
Microsoft Kicks Unsupported PCs From Windows 11 Testing Program

As the October launch date for Windows 11 grows closer, Microsoft is standing firm on the new software's spec requirements. In fact, anyone running the beta on unacceptable hardware is being booted from the program, and their only option is to accept defeat and reinstall Windows 10.

Google’s AutoML Creates Machine Learning Models Without Programming Experience
Google’s AutoML Creates Machine Learning Models Without Programming Experience

The gist of Cloud AutoML is that almost anyone can bring a catalog of images, import tags for the images, and create a functional machine learning model based on that.