Microsoft Teams With SpaceX To Push Cloud Battle With Amazon Into Orbit (zdnet.com) 31
Microsoft is teaming with Elon Musk's SpaceX and others as the software giant opens a new front in its cloud-computing battle with Amazon.com targeting space customers. From a report: Microsoft would help connect and deploy new services using swarms of low-orbit spacecraft being proposed by SpaceX [Editor's note: the link may be paywalled; alternative source], and more traditional fleets of satellites circling the earth at higher altitudes. Microsoft's initiative targeting commercial and government space businesses, formally launched Tuesday, comes about three months after Amazon Web Services, the e-retailer's cloud unit, disclosed its space-focused effort. Some analysts have projected that overall revenue from space-related cloud services could total about $15 billion by the end of the decade, at least several times higher than current levels. Competition in the cloud between Amazon, the market leader, and No. 2 Microsoft has been heating up in recent years. The pandemic has intensified the fight as companies accelerate their shift to the cloud and make vendor choices that could last for years. [...] SpaceX, which is in the process of deploying its Starlink project consisting of thousands of high-speed internet satellites intended to provide connectivity around the globe, makes a natural partner for Microsoft. A major reason is that Amazon founder Jeff Bezos is pursuing his own low-orbit satellite constellation. Mr. Bezos also owns Blue Origin, a rocket company competing with SpaceX.
"Microsoft Teams with [insert name here]" (Score:4, Insightful)
Microsoft Teams with [insert name here] usually doesnt end all that well for the partner company - good luck there Space-X.
Re: (Score:2)
There used to be some great lists of former Microsoft partners online. Does anyone have a link to any of those? They are really hard to search for with Google recently.
For the next Blue Screen of Death (Score:2)
We'll get hit by a toilet seat from space?
Re: (Score:2)
Edge computing (Score:4, Informative)
There's more to it, but that's the basic premise. They believe (and not without reason) that StarLink is going to become a major player and they want to be positioned (literally and figuratively) to provide the fastest-possible access through same to their services.
Not the worst idea out there.
At first, it sounded like they were going to put Azure data centers in orbit, such that StarLink wouldn't even have to hit a ground station to hit them. Didn't seem like a good idea, as such systems would be hideously expensive to put up there. That is NOT what this is.
Re: (Score:2)
Thanks for clearing that up.
I was wondering what kind of processing power they could get into space myself. I mean, running just a quiescent windows 10 takes a bit of power to idle on. Then the whole First Tuesday of the month patch update thing would bring star link to it's knees... A hundred big blue marble screens of death, with the "any key" you need to press to continue being in orbit.
Blue vaporware (Score:4, Interesting)
Wake me up when Blue Origin actually puts something in orbit.
Re: (Score:2)
Wake me up when Blue Origin actually puts something in orbit.
No, actually you don't want them to also put shit in orbit.
None of us do.
The last place we need capitalistic cutthroat competition, is in the already crowded orbits around our planet.
This will become very clear to the GPS Generation of Social Media junkies when Facebook goes offline permanently, and they struggle to navigate their way out of the bathroom without google maps guiding them to a roll of toilet paper.
Re: (Score:2)
OK, for one thing, space is big. Really big. A whole lot bigger than the surface of this planet that we're used to basing our assumptions on. So while we do need to manage space junk, there's a huge amount of room left for legitimate satellites.
Secondly, nations (US, India, China) have already started testing destruction of satellites and other aggressive moves. So we have already failed at keeping competition out of space. Can commercial competition be far behind?
Thirdly, you're absolutely right about soci
Re: (Score:2)
OK, for one thing, space is big. Really big. A whole lot bigger than the surface of this planet that we're used to basing our assumptions on. So while we do need to manage space junk, there's a huge amount of room left for legitimate satellites. Secondly, nations (US, India, China) have already started testing destruction of satellites and other aggressive moves. So we have already failed at keeping competition out of space. Can commercial competition be far behind? Thirdly, you're absolutely right about social media junkies, but I'm afraid FB, Google, et. al. have already got too much of a stranglehold on our culture, and by extension our government, to be legislated or sued out of existence. That horse has already left the barn.
Twenty years ago, NORAD/USSPACECOM was tracking something like 16,000 objects in orbit. Everything from used boosters to a wrench. And when we're talking about particles traveling at 18,000MPH able to rip through a fuselage/spacesuit/human like a hot knife through butter, it's a fucking problem.
And just because idiots are wanting to take our warmongering into orbit doesn't mean those "aggressive" idiots are any smarter than the "aggressive" idiots fighting 5,000 years ago on the ground. You think other
Re: (Score:2)
Twenty years ago, NORAD/USSPACECOM was tracking something like 16,000 objects in orbit. Everything from used boosters to a wrench. And when we're talking about particles traveling at 18,000MPH able to rip through a fuselage/spacesuit/human like a hot knife through butter, it's a fucking problem.
Don't look now, but they were tracking over 500,000 [nasa.gov] objects in 2013. There are many more now. And yet all of the orbital planes remain quite usable. Funny, that. Probably because everything in low Earth orbit is moving at around the same speed as everything near it. Because that's how orbits work, and nobody launches things into retrograde orbits except for Israel [wikipedia.org], who has successfully launched 10, six of which are still in orbit, and the US, who have launched two [wikipedia.org], and everybody stays the hell away fro
Re:Blue vaporware (Score:4, Funny)
SpaceX would like to know more about your long-term cryogenic suspension technology.
Re: (Score:2)
Wake me up when Blue Origin actually puts something in orbit.
SpaceX would like to know more about your long-term cryogenic suspension technology.
Liquid nitrogen burn...
We don't need an other Captain Midnight take HBO (Score:2)
We don't need an other Captain Midnight like thing take HBO Down For windows updates.
What is this? (Score:3)
I can see why a containerized datacenter in a remote location would need bandwidth that StarLink can provide. I am having more trouble imagining what kind of remote datacenter would need a small enough amount of bandwidth that satellites could provide it. The ratio of bandwidth to computation of that combination is minuscule. And if satellite bandwidth is decent, why have a datacenter onsite in a remote location at all?
Maybe some high-bandwidth local datalinks and sensors that need heavy-duty signal analysis and reduction before they can uploaded via satellite?
I'm not denying this new market exists, I just want a concrete example of where it would be workable and beneficial.
Re: (Score:3)
I imagine they will co-locate some Starlink base stations and Microsoft cloud servers, for reduced latency.
Putting servers on satellites probably wouldn't be a good idea due to power requirements for anything significant.
Re: (Score:2)
...I can see why a containerized datacenter in a remote location would need bandwidth that StarLink can provide. I am having more trouble imagining what kind of remote datacenter would need a small enough amount of bandwidth that satellites could provide it...
Spent a few years deploying tactical LAN/WAN solutions. Back in the 90s we had a single T1 satellite shot to feed about 200 users on the field. Email, internet, secure comms, etc.
Regardless of what your local CDN Data Pimp wants to sell you, the world does not require a dozen fiber lines worth of bandwidth and 4K streaming to actually work and communicate. I can still fit a few dozen text files in 1.44MB worth of space today. Basic communication and functionality, did not change.
Re: (Score:2)
I suppose the question would be what a cloud DC with a satellite uplink would provide better than a cloud DC with a traditional uplink. For all but starlink clients the DC would be worse connectivity, and for Starlink clients the benefits would be miniscule compared to going a bit further.
I can see edge deployments that would benefit from satellite uplink (performing some autonomous activity at a site that needs to be closed loop for internet outages or to minimize traffic), but a hosting solution does not
Re: (Score:2)
I can see edge deployments that would benefit from satellite uplink (performing some autonomous activity at a site that needs to be closed loop for internet outages or to minimize traffic), but a hosting solution does not seem to benefit from being out in the middle of nowhere...Unless this is some sort of variation on a theme to say you have an 'on-premise' solution that is, for no good technical reason, hard-linked to one particular cloud provider versus another.
Ironically enough, it only took me about half a second to think of one variation on a theme that existed before:
https://en.wikipedia.org/wiki/... [wikipedia.org]
And with mega-corps catering to government overreach (basically becoming illegal extensions of our intelligence communities), I would imagine the demand for those kinds of "variations", are going to grow, and continue to carve out digital Switzerlands.
Re: (Score:2)
Sure, but the question is why would you need a shipping-container's worth of processing to handle that data flow?
For what you are describing, a portable mobile cell tower with a StarLink uplink seems to
Re: (Score:2)
Sure, but the question is why would you need a shipping-container's worth of processing to handle that data flow?
Probably for the same reason I found a personally-owned 40-pound NAS array rack-mounted in a transit case before a deployment to the middle-o-nowhere.
Ain't nobody got time to download porn on a T1 line.
Re: (Score:2)
Had to deploy a Pelco Endura security video solution at the far end of a T-1 at one time (telecom delayed its planned fiber installation for two years so it's what we were stuck with). That was a miserable experience, forget reliably launching video feeds, we couldn't even maintain connectivity with the management tools over that link.
Re: (Score:2)
I'm just wondering what is "new" about the Azure Modular Datacenter (MDC). I set up security for a MS data center around 2008 that we just called "the trailer park" because it was a bunch of containerized DC modules surrounded by a fence with a couple of double-wides for the admins and data techs to hang out in. IIRC they had bought the containers pre-packed with pre-imaged servers from Dell, who had developed the product to sell to disaster recovery companies. Hell, several years ago they dropped one in
Re: (Score:2)
I can see why a containerized datacenter in a remote location would need bandwidth that StarLink can provide. I am having more trouble imagining what kind of remote datacenter would need a small enough amount of bandwidth that satellites could provide it.
You haven't been paying attention. SpaceX demonstrated Starlink [reuters.com] precursor technology with the US Air Force achieving 610 megabits per second to an aircraft in flight. Bandwidth to a fixed ground installation is expected to be even higher, and for a special customer like Microsoft, they may be willing to devote the entire transceiver bandwidth of the satellite that's overhead. Estimated max throughput for the entire satellite is around 20 gigabits per second. That's pretty decent if you're the lone custo
Re: (Score:1)
Except it's really nowhere close to 20 Gbps, and also, the 610 Mbps demonstration used a Ball Aerospace (not Starlink) antenna and an entire satellite was dedicated to it.
Starlink's current capability is closer to 3 Gbps per satellite if using their terminal and each one costs about $1 MM to make and launch. Actually, their terminals use a bit under 250 MHz of bandwidth, so, their theoretical top speed is a couple hundred Mbps, which has to be shared between uplink and downlink. Not saying you can't connect
Weird pedantry (sorry) (Score:3)
I hate to be weirdly pedantic, but can Slashdot stop this abusive capitalization?
"Microsoft Teams With SpaceX To Push Cloud Battle With Amazon Into Orbit"
I was trying to figure out how Microsoft Teams was going to work with SpaceX to create a Cloud Battle with Amazon in some sort of stadium called "Orbit". Given that MS uses Teams as a brand, this was super confusing capitalization.
Too many small things... (Score:2)
No. Just No. (Score:1)
There is no compelling argument for putting these resources in LEO. None. Starlink is building the network, leave the damned servers/etc on earth.
We do not need more objects in LEO or anywheres near that. The risks of collision and damage only increase, and it is time to consider that while space 'is big, really big', it is not infinite, and congestion is a real probability. Let's put that off until we have the technology, resources, and/or will to deal with that problem. We do not have all three now.
Do you want Skynet? (Score:2)
Because this... this is how you get Skynet.