Starlink, the satellite Internet provider from Elon Musk's SpaceX, will be able to support just 485,000 simultaneous users at 100Mbit/s across the entire US, according to one firm's new estimates. And that kind of performance won't even be available until the end of 2026, when Starlink floods Earth's skies with up to 12,000 satellites.

According to the new estimates from the financial analysts at Cowen, those figures mean that Starlink – a low Earth orbit (LEO) satellite Internet operator that is expected to begin offering a public beta service in November – won't pose much of a threat to established Internet service providers like Verizon and Comcast.

"While Starlink has the ability to provide a practical satellite-based broadband solution for the underserved, the capacity has limitations in most of the US especially considering the growing demand for bandwidth driven by in-home data-rich applications and devices," the firm wrote in a new report Wednesday.

The Cowen analysts said they developed their Starlink model based on a wide range of public statements – including in FCC filings, Tweets and other sources – by the private company and its executives over the past several months. Nonetheless, the firm's findings offer perhaps the most cohesive view yet of Starlink's ambitions and obstacles.

As the Cowen analysts note, Starlink today operates around 650 satellites in orbit – an impressive feat considering there are so far only a total of 3,000 satellites in operation today across the entire world. Moreover, Starlink has plans to dramatically increase its satellite fleet to 4,400 by 2024 and, potentially, 12,000 by 2026.

Starlink's system will work much like those from other, existing satellite Internet providers: Customers with a small Starlink antenna will beam their signal to a Starlink LEO satellite, which will bounce that signal down to a nearby ground station that will connect the customer to the Internet. However, the firm notes that Starlink itself has told the FCC that "each satellite in the SpaceX system provides aggregate downlink capacity to users ranging from 17 to 23Gbit/s."

"Thus, assuming 100% efficiency (not realistic, but we are simply providing context as a high book-end), and assuming 20Gbit/s per satellite implies that each satellite can handle 200 simultaneous streams at 100Mbit/s," the analysts wrote.

In crunching the number of satellites that would be covering the US at any one time, the analysts conclude that Starlink can serve 485,000 simultaneous data streams in the US with 100Mbit/s speeds if all 12,000 Starlink satellites are operational.

However, they noted that this figure doesn't directly correlate with the number of actual customers Starlink can support, because not all customers will be online at the same time. This concept – called "oversubscription" – is a common calculation for network operators to make, whether they're 5G providers or cable companies. They must decide how many customers to allow onto their network, knowing that only a certain percentage will be using it simultaneously at any one time.

"Equating this number to actual customers depends on Starlink's policy on oversubscription rate, for example assuming a 3x oversubscription rate could imply a 1.5 million total addressable market (not all customers will be online simultaneously, far from it, though we believe the post-COVID era will drive more need for Internet reliability, thus more conservative oversubscription levels)," they wrote.

Finally, in their report on the topic the Cowen analysts argued that Internet users continue to consume more and more data every year – and they said they expect that usage to continue to rise amid pandemic-era work- and school-from-home lifestyle changes.

"We contend that US broadband consumption, and the speeds that users demand, is continuously growing," they wrote. "Thus, as satellite throughput and technology continues to progress, so too will demand for faster speeds. As such, our analysis shows that LEO satellites will continuously be a step behind wireline telco/cable operators in meeting US consumer demand for broadband."

None of these calculations would likely come as a surprise to Starlink backer Elon Musk.

"Starlink is not some huge threat to telcos. I want to be super clear: it is not," Musk said earlier this year at a trade show. "5G is great for high density situations, but it's actually not great for the countryside, you know, for rural areas. It's not great; you need range. And so in any kind of sparse environment 5G is really not well suited."

— Mike Dano, Editorial Director, 5G & Mobile Strategies, Light Reading | @mikeddano