Connect with us

News

SpaceX Starlink partners with Microsoft Azure to deploy cloud computing anywhere

Published

on

Microsoft Azure has announced a partnership with SpaceX that will give customers the ability to both access and deploy cloud computing capabilities anywhere on Earth with the help of Starlink internet.

Ultimately designed with anywhere from ~4,400 to ~40,000 operational satellites in mind, SpaceX’s Starlink constellation aims to connect users to the internet where existing access is either too expensive, limited, or completely unavailable. Of course, however, connecting the world’s unconnected is an immense and challenging aspiration – one that is unlikely to be one of the Starlink constellation’s first major uses.

As CEO Elon Musk has been keen to regularly note, the real challenge of SpaceX’s Starlink satellite internet project is ultimately ensuring that the constellation doesn’t join the graveyard of bankrupt companies that came before it. For better or worse, that will necessitate close relationships with as many premium enterprise-class customers as possible. With its estimated 2020 market cap of ~$370 billion expected to grow to ~$800 billion or more by 2025, cloud computing is one such potentially lucrative application.

To better exploit the benefits offered by the kind of blanket connectivity Starlink may soon offer, Microsoft has developed its own Azure Modular Datacenter (MDC), essentially a data center built into a mobile, satellite-connected shipping container. Customers can choose to either use the MDC as a wholly independent datacenter or connect it to one or more satellite constellations, Starlink included. With what a SpaceX executive recently described as dual parabolic antennas, an MDC could likely have access to gigabit-class internet connectivity with latency comparable to fiber anywhere on Earth.

According to Microsoft, possible scenarios where an MDC would be valuable include “mobile command centers, humanitarian assistance, military mission needs, mineral exploration, and other use cases requiring high intensity, secure computing.” Several Azure Mobile Datacenters have already been deployed and are being trialed by private sector companies and the US military.

Likely less than coincidental, Microsoft Azure’s Starlink partnership comes around the same time as Amazon has begun to peel back the curtains on Project Kuiper, a low Earth orbit (LEO) satellite internet constellation almost indistinguishable from Starlink. Lead and largely staffed by former Starlink executives and employees, Project Kuiper aims to deploy a constellation of ~3200 small, interlinked communications satellites – a goal Amazon has pledged at least $10 billion to achieve.

Advertisement
-

Somewhat unsurprisingly, Kuiper – lead by executives SpaceX CEO Elon Musk personally fired in 2019 for moving too slowly – has no set schedule or indication of early prototype development and is effectively 3-5 years behind SpaceX, OneWeb, and other prospective constellation operators from the get-go.

SpaceX has launched 773 operational Starlink satellites in less than one year. (SpaceX/Richard Angle)

An IEEE Spectrum article offers an excellent summary of the web services. logistics, and online shopping giant’s most likely motivation behind investing so much money in a satellite constellation that is – at best – years behind.

“‘With Amazon, it’s a whole different ballgame,’ says Zac Manchester, an assistant professor of aeronautics and astronautics at Stanford University. ‘The thing that makes Amazon different from SpaceX and OneWeb is they have so much other stuff going for them.’ If Kuiper succeeds, Amazon can not only offer global satellite broadband access—it can include that access as part of its Amazon Web Services (AWS), which already offers resources for cloud computing, machine learning, data analytics, and more.”

Michael Koziol – IEEE Spectrum – 17 August 2020

In other words, Amazon likely believes that its potential advantages are so strong and so unmatched that it doesn’t matter if it’s years late to the party. On the other hand, it could also be the case that Amazon – and Amazon Web Services in particular – perceives a lack of the capabilities offered by a high-bandwidth satellite internet constellation to be such an existential threat that the company has no choice but to try to enter the fray.

As such, SpaceX’s partnership with Microsoft Azure Cloud Services is a direct shot across Amazon’s bow, demonstrating that even if Project Kuiper manages to begin operational satellite launches in just a year or two, the company will immediately face experienced, organized competition. There is some level of irony in the fact that, purely out of corporate spite, Amazon will now likely never become a Starlink customer to avoid helping a direct competitor, meaning that AWS will be consciously putting itself at a competitive disadvantage for years to come by waiting for Project Kuiper.

Eric Ralph is Teslarati's senior spaceflight reporter and has been covering the industry in some capacity for almost half a decade, largely spurred in 2016 by a trip to Mexico to watch Elon Musk reveal SpaceX's plans for Mars in person. Aside from spreading interest and excitement about spaceflight far and wide, his primary goal is to cover humanity's ongoing efforts to expand beyond Earth to the Moon, Mars, and elsewhere.

Advertisement
Comments

Elon Musk

Google just picked SpaceX for its first step into orbital AI

Google will launch its first Project Suncatcher AI satellite on SpaceX’s Transporter-18 rideshare next week.

Published

on

By

Google is about to put its own AI chips into orbit for the first time, and it is paying SpaceX to get them there.

The company said Thursday that the first in-orbit test of Project Suncatcher, its research effort to find out whether space can host large-scale AI computing, will fly next week on SpaceX’s Transporter-18 rideshare mission.

The satellite, called MVP, is about the size of a refrigerator and carries four of Google’s Tensor Processing Units, the same chips Google runs in its ground data centers. Google originally planned to launch two custom satellites in 2027, but chose to move faster by integrating its chips into a satellite.

MVP’s solar panels supply about one kilowatt of power, and Google will run Gemini models on the TPUs only in bursts of roughly 15 minutes before the chips shut down so the radiators can shed heat. In a blog post, Google said its Trillium TPUs survived vibration testing that mimicked sustained launch loads of up to 10g, with individual components seeing 50 to 100g, and handled a radiation dose greater than a five year mission would deliver.

SpaceX and Google mull massive partnership on Musk’s orbital data dream: report

Next week’s flight, slated for October 1, follows a relationship that became public in May, when Teslarati reported that Google was in talks with SpaceX for a launch deal tied to orbital data centers. Google also holds a stake of roughly 6% in SpaceX.

The two companies are chasing the same idea from very different starting points. SpaceX’s own orbital compute program is built around the AI1 satellite, a roughly 70 meter structure derived from Starlink V3 hardware that is designed for 150 kW of peak compute, about 150 times the power MVP will draw. Elon Musk has brushed off concerns about crowding orbit with those satellites, and SpaceX is building its Gigasat factory in Bastrop, Texas, to produce them, targeting an annualized rate of about 1 GW of space compute by the end of 2027.

Musk also posted on X on Thursday that “the amount of compute in space will obviously round up to 100% of all compute.”

Google has been more cautious in public. Its research estimates that launch prices need to fall below about $200 per kilogram before an orbital data center can compete with a ground facility on energy cost, a threshold the company believes could be reached around the mid 2030s. The Suncatcher team has said it expects the effort to remain a project rather than a product for years, which leaves the first real test of its hardware riding on a rocket from the company with the most aggressive timeline in the field.

Continue Reading

Elon Musk

Tesla Cybercab gets initial tie-in to localized, in-house cathode plant

Published

on

Credit: Tesla

Tesla has taken another concrete step toward owning its battery supply chain, and it’s doing so with what is perhaps the most important vehicle in its short-but-storied history.

On September 23, Tesla announced that it has officially built the first Cybercab with cathode material produced in-house at the company’s first cathode plant in the U.S., and the first in the U.S. overall.

Active cathode material is the most expensive piece of a lithium-ion battery cell, and it often accounts for more than a third of cell cost. For years, the industry sourced a majority of it from Asia, but Tesla’s decision to make it in the United States bodes well for the Cybercab project. This is the latest chapter in Tesla’s vertical integration strategy, which began in public at Battery Day in 2020.

At the Battery Day Event, Elon Musk said the company would build a North American cathode plant and overhaul the process to cut costs and waste, while also making some of the most powerful and long-lasting cells in the industry.

The Austin facility took years to appear. Tesla filed permits for “Project Cathode” in 2022 on land near Giga Texas. By mid-2022, the building frame was up and Tesla later invested hundreds of millions of dollars as part of a larger expansion of the Giga Texas plant. The company stated it was operating the first large-scale cathode production facility in North America to supplement 4680 cell production.

One month later, that material reached a finished Cybercab.

The timing of this breakthrough is monumental for the Cybercab program. As Tesla officially launched the first Cybercab rides to the public earlier this month, production of the ride-hailing-geared vehicle is moving forward on the planned S-curve that CEO Elon Musk told everyone to expect.

Nevertheless, packs of Cybercab units have been spotted throughout the United States, in an effort to potentially activate the fleet as soon as the company gains regulatory approval in various geographic areas.

On top of that, Tesla owning the cathode step and pairing it with its own in-house lithium from the Gulf Coast refinery shortens the supply chain that once stretched thousands of miles and subjects every pack to fewer external price shocks and geopolitical risks.

Tesla is not yet independent of all of its foreign suppliers, as some precursor metals come from mines and chemical plants. But the first in-house cathode Cybercab shows the company is closing the most expensive and most concentrated gap in its battery production efforts. For a vehicle like Cybercab to operate at a high utilization within the Robotaxi network, that control over cost is so crucial.

It is arguably as important as the software that drives it.

Continue Reading

News

Tesla leaks Semi customers ahead of handover event

Published

on

Credit: Tesla

Tesla is set to inaugurate its dedicated Semi factory in Sparks, Nevada, this evening, but drone footage has already started to show who is expected on the handover stage.

Zanegler, a drone operator who regularly documents Tesla Semi operations in Nevada, filmed a row of new-generation Semi units parked outside of the plant with customer branding applied. The liveries visible in the video include companies like PepsiCo, U.S. Foods, Einride, DHL, WattEV, and others that have committed to utilizing Tesla’s Class 8 electric truck in their fleets.

Several of those company names already sit within Tesla’s public ordering book, including PepsiCo, which has been the program’s primary anchor customer since the first Semi deliveries in December 2022. PepsiCo now runs dozens of Tesla Semi units in daily service, with drivers completing regional routes.

Einride placed a 500-unit order in August, which was, at the time, the largest single Semi commitment. First deployments of their Semi units are planned for this month and will eventually span across California, Texas, New Jersey, Illinois, and Georgia. Meanwhile, IMC Logistics announced a 50-truck California order this week.

Tesla Semi lands the biggest electric truck deal in U.S. history

Other names, like LTS, OK Produce, and HMD, have not been widely confirmed as firm buyers, but their presence on the lot is perhaps the clearest signal that Tesla intends to expand on that list this evening at the event.

Tesla first unveiled the Semi back in 2017 with a 2019 production target.

Limited pilot builds reached PepsiCo in late 2022, and high-volume production began at the new 1.7-million-square-foot plant that sits logically positioned next to Tesla’s Gigafactory Nevada. Tesla Semi program director Dan Priestley has said that Tesla expects to build “many thousands” of Semi units this year, but as with every Tesla release, there have been set expectations of an S-curve ramp.

Outside estimates still put 2026 deliveries in the 5,000-to-15,000 range, but it will ultimately depend on Tesla’s readiness to put those units out, as well as how smoothly production is moving internally.

Tesla’s Semi Rollout event is set to start tonight at the Sparks factory, and it will be livestreamed on X starting at 6 p.m. PDT / 9 p.m. ET.

Continue Reading