Wouldn't the better comparison be if they just rented the planes from the manufacturer though? Or even better, have a 3rd party company buy and maintain the aircraft and they just rented the aircraft from them. What could go wrong?
It absolutely is not as simple as you seem to make it sound though... a poorly experienced team will end up costing more in time and money than the markup of an outsourced but scaled efficient team.
The US government outsources everything. Anything actually being produced or operated is done by some vendor on a government contract, with the big exception of the military (which still has plenty of outsourced business).
Government IT expertise is not high-quality and will never match what AWS has. It's about value, not cost. It's better to have reliable and scalable infrastructure even if AWS is vastly more expensive.
It just takes a lot of people and experience to do that. There are a few companies out there that can are already set up well for that. I don't know if I'd put Oracle in that list but Amazon isn't the only one with the experience.
At US government scale, they’d come out ahead if executed by the US Digital Service (no politics, executed rapidly). Amazon has outrageous margins on AWS, margins that could be better spent by the government on equipment and technical talent.
Ideally, they’d contribute to OpenStack through such an initiative, ensuring an open source alternative to commercial closed cloud providers.
I don’t want my tax dollars going towards Bezo’s attempts at a retail and cloud computing monopoly, but that’s just me. Maybe there’s an opportunity here for Trump considering his disdain for Bezos and the USDS being part of the Executive Office of the President.
There is way more that goes in to operating a large scale computing infrastructure efficiently than the cost of hardware. Like way, way more. If anyone could replicate it at the right scale then every large organization would be competing with Amazon for the cloud infrastructure market.
I ran a couple data centers on either coast each burning about half a megawatt, which isn't that much in the grand scheme of thing. My employer at the time offered a SaaS product to Fortune 100 companies. The product required a realtime kernel and QOS in the network, neither of which is supported in AWS.
I did a fair bit of math comparing costs. Typically, you could pay off the CapEx with 6 months of usage in AWS.
IIRC we had around $8m in salaries for the ops org. I'm not sure we would've needed a great deal fewer people having all that infra in AWS.
The real problem comes from trying to find a team of people willing to engineer your service for failure. That kind of attitude just doesn't exist in the world of IT, so anyone would be hard pressed to put a team together.
Good luck finding a network engineering team that knows BGP well enough for such a thing. Good luck finding a team of engineers that know Openstack well enough to keep it running. Good luck finding folks to run Ceph. Good luck getting all that right.
If you could, though, there are certainly savings to be had.
We aren't talking about a datacenter to support one product. We are talking about an entire platform for building and operating any number of products on top. It isn't comparable.
Can you prove your case? If you are correct it's just a matter of scale, it stands to reason that any large organization would be muscling in on the high-margin cloud computing business. That clearly is not the case. But I'm happy to see any evidence you have to the contrary.
Sometimes there are diseconomies of scale as well. Like all of a sudden you have to put a data center into an otherwise shitty location because the committee chairman needs to show some results to constituents.
Or two departments have essentially the same requirements, but they compete for budget and each have to justify their staffing by duplicating effort.
(These pathologies aren't unique to government. You can also find them in large private companies.)
There's a certain scale where it is definitely possible to do it cheaper, but the government isn't really able to do that. Even if the program didn't crash and burn (fairly likely- there's plenty of examples of billion dollar government software programs that turned into vaporware), it'd be taking on a ton of risk just to set up the same thing without the AWS profit margin (maybe 10-20%?)
Just because they spent $20 billion on this, and $20 billion on that, doesn't mean that they also have another $20 billion they can spend on something else as well. There's no logic in that argument.
Large organizations have the cash flows and reserves to realize efficiencies smaller organizations cannot.
You cannot afford to own a refinery. Delta Airlines can and does, as a hedge against fuel costs (which turned out to be a very profitable decision for them). That is the point.
I never understand why the us government wants to contract these things.