From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-oa1-x30.google.com (mail-oa1-x30.google.com [IPv6:2001:4860:4864:20::30]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id AAB993B29E for ; Thu, 6 Feb 2025 10:29:31 -0500 (EST) Received: by mail-oa1-x30.google.com with SMTP id 586e51a60fabf-2b832eab8cdso163885fac.0 for ; Thu, 06 Feb 2025 07:29:31 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1738855771; x=1739460571; darn=lists.bufferbloat.net; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:from:to:cc:subject:date :message-id:reply-to; bh=3i4Rp5+19VkVxUkjMTViNYyufh4mWs5Lg56eSc6csWU=; b=heVY0g0PPkrLE07qaSwQrJUasdEi99Ms8tQgxnWZjagQBi4i+bNcPj8o/GezRXgVi2 3RzLsGIyQWcL8JnMpeXfjWLnfRAsqPA+7galR+Stgjeqo7TPrvc47f4ZuDjWF/lNDkjw wTeEvQnAhnAs2baERcz+vWo1q76ihn25m6pdcVP8Oev3YV2/Boy0sIbwnWkG95EdTPpB b5z2GtiBUMmaDTopS07SaZwrlZzyYCrB+WKA+iJmxuT11QFzEQVy4nuCwM78xNTzFFzD dAiagUTFWN2nFLQ9srWD/ZrZkCyYzWQ7Ytw+VsVTLzNUE6Bmc2Xb/cGg8YZJWZ6FB5r5 TQpw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1738855771; x=1739460571; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=3i4Rp5+19VkVxUkjMTViNYyufh4mWs5Lg56eSc6csWU=; b=SDdffKOMJFBDmiTEHGj2f/iiBTMz8u/+YMYlk4s5IjH7jrO+d33pQ0ywDdijn14V7J V+uPNJoaWvQxQo96/HTFx2/vOM8hv5oiftsb6R6ayAH49+CawhayVuZZG7hJhonNwSzV o2tdmmIAGbXLdB5Ct4BXdKpmlOqQp1ehJygljVeh6owWJV9hPWq87roh56/nB2kx+sZH NoG4Xlt9Kji/ItdgexH53qOdXByXkQu6xD/H2PNoNthXQX2gduNI+hrhTvOg7Nw4Dvep 65WoeTGSTLXUqikqVuB7nB67AF+wGBRxbXbKC/IQebXto3Y1rmO/tGt419og7fzGcA/v qlIw== X-Gm-Message-State: AOJu0YwpAaV1+0BgFT0KGZu5XGf285dZHA0OzilouXcP4Yh1Erg8F5gw iqVtOAPkrN/w+placThZU2r3xKphiWNRq+wstVuFO23o7qFF4BeDEnZ62Hd5CpQYtJHBDdLod0I g6mJXpyXC2B/2UfzUOfaNu2khDw3uTqqd X-Gm-Gg: ASbGncsoBQMYQ5v1d959Ox5/q3ba/24/hyR/kdVAt8IbeN4N44EjraG3XUiDjIBpCS4 dghr3m/T2uFa02dv2zDjKuHRpNmqGvQc9Os+ATMAjlFNu0oNpm1ZHtzDAcBagVRrup3YSLcqz++ mjtVSNWyjayfqRixAI6PtZaZLWcg73Sk4= X-Google-Smtp-Source: AGHT+IH9mTcBaPmxgWSHwI8KzeWrQRDwlBv0bEjfKDSVjZG05TsYlWt17jhfAkW9xYSNVrj75XA8L/RAveLFmTn/TG0= X-Received: by 2002:a05:6870:c0cc:b0:29e:76c8:be2e with SMTP id 586e51a60fabf-2b8050bb69bmr5373490fac.28.1738855770885; Thu, 06 Feb 2025 07:29:30 -0800 (PST) MIME-Version: 1.0 References: <4cadae1b-e31e-4dae-92c1-9d847edf7489@auckland.ac.nz> In-Reply-To: <4cadae1b-e31e-4dae-92c1-9d847edf7489@auckland.ac.nz> From: Dave Taht Date: Thu, 6 Feb 2025 07:29:18 -0800 X-Gm-Features: AWEUYZldrkmM1vuoX6svmT7r9WC43bUgGX2kmo1546wKRyuERnhgwjaH_79jCZQ Message-ID: To: Ulrich Speidel Cc: starlink@lists.bufferbloat.net Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Subject: Re: [Starlink] Starlink profit growing rapidly as it faces a moment of promise and peril (Ars Technica) X-BeenThere: starlink@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: "Starlink has bufferbloat. Bad." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 06 Feb 2025 15:29:31 -0000 On Thu, Feb 6, 2025 at 12:47=E2=80=AFAM Ulrich Speidel via Starlink wrote: > > Now the interesting thing here is that with 5 million subscribers paying = about US$1200 a year, you'd get about 6 billion from bog standard dishy end= users alone. So that $8.2b is credible. > > Note this is revenue, not profit. To get there, pointing at the other pos= ts today, Starlink had to build a constellation of about 7,000 satellites. = Even if we looked just at these 7,000 and assumed incorrectly that they all= got to enjoy a full service life of maybe 5 years, As best as I remember from something a few years ago, the fuel usage was far, far below expectations, something like 3% over 3 years. So aside from the lowest orbits perhaps, starlink does not need to retire a satellite in under 5 years, except for upgrades. > we'd be looking at 1,400 of them needing to get replaced each year going = forward. They are presently launching 22 sats every other day. call it 4000/yr. So assuming the same launch cadence, call it 21k sats in orbit before they can grow no more. >Assuming here 1000 kg per satellite going forward (just ballpark) and US$1= 000/kg launch cost. So that's a US$1M replacement cost per satellite (not e= ven looking at the hardware), and that's got to come out of those $8.2b. This seems to assume each launch is 22m. I have no idea what the second stage costs at this point... >So I guess profit might be closer to the $6b mark at best in that scenario= , and probably nowhere near that so far due to the fact that SpaceX are lau= nching at well beyond replacement rate, the launch costs of anything older = than Starship are higher, and the V3's will be closer to 2000 kg than 1000 = kg. So that mightn't leave quite that much change out of the $8.2b to throw= at other projects. But it's certainly looking like a sustainable business. Actually connecting to the internet costs money too, and each ground station is probably 1m+, and there are all sorts of other costs in manufacturing and network support. Still it certainly looks like a sustainable business. > > But then Starlink are growing. I guess with Elon now being in de-facto co= ntrol of the FCC, they'll get what they want, but more sats up there also m= eans having to replace more eventually. So that cost will go up. The FCC has lost authority to regulate the internet via title II, which is good because they had no staff for it. Historically Brendan Carr (and Nathan Simington and Ajit Pai) have been quite friendly to all forms of wireless, be it 5g, or wisps, or sat. So while I expect a friendly spectrum regime, not much on the internet front. What happens next at the NTIA is anyone=C2=B4s guess. > > It then depends on revenue growth, and that in turn depends on: > > capacity available to sell and > markets to sell into. Some very large countries like India, are dragging their feet. I do wonder what will happen in onterio, where they just uncanceled a 100m starlink contract ( https://www.reuters.com/business/media-telecom/ontario-cancels-starlink-con= tract-latest-canadian-tariffs-protest-2025-02-03/) and are trying to put 2.5B into another sat system. Once the furor dies down over there, I have no idea what will happen. In general I expect most future growth for starlink to be in rural throughout the world, and their initial estimates were for 30m users worldwide, which I think is quite achievable in 5-7 years. > And here lies the crux: Capacity comes in two types: > > Spectral capacity. That's SpaceX's ability to find a frequency to serve a= customer on that isn't already in use in the customer's neighbourhood. > Beam capacity: The ability to find a spare beam on a satellite that can b= e used for that customer. > > Now the second one of these is easy to address - just launch more sats an= d put more beams on each sat. But the second capacity is worth nothing in a= place where you don't have the first one, which can't be increased by laun= ching more satellites - at least not unless they're different ones that all= ow for smaller cells and sharper beams. That's a route that Starlink are tr= ying to go down FAIK, but there's limited growth potential here. > > And looking at the Starlink availability map, spectral capacity is someth= ing they currently seem to be grappling with in quite a number of places. F= rom the Greenwich Meridian roughly east, they're "sold out" in: the greater= London area, Accra, Lagos, Benin City, Warri, Port Harcourt, Abuja, Lusaka= , Bulawayo, Harare, Maputo, Nairobi, Antananarivo, Jakarta, Perth, Manila, = Brisbane, Bethel, quite a lot of areas south of Anchorage, spots around Fai= rbanks, Delta Junction, Whitehorse, the Seattle-Portland corridor, Sacramen= to, Grande Prairie, Spokane, San Diego, Missoula, Edmonton, Apache Junction= , Nogales, Aspen Park, Guadalajara (MX), Monterrey (MX), Mexico City, Austi= n, Puero Escondido (MX), De Ridder, M=C3=A9rida (MX), San Salvador, Playa d= el Carmen (MX), Peterborough (CA), Tuskegee, San Jose (CR), Highlands (NC),= much of western Jamaica, parts of the Dominican Republic, much of Puerto R= ico, Iqualuit, Leticia (BR), Rinc=C3=B3n de Los Sauces, Sao Gabriel da Caoc= hoeira, Tef=C3=A9 (BR), Manaus (BR), Sao Paulo BR), Rio de Janeiro. It's be= en like this for a couple of months now, so I guess it's not a problem with= the Dishy supply chain. That=C2=B4s a really great list. However, in terms of square footage quite = small. > Read: Not much growth potential at present in and around population centr= es where Starlink used to be available and where there isn't good existing = ground infrastructure. > > There used to be a lot of availability "flickering" in areas where there = was more demand than beam capacity - this has gone solidly to "available" n= ow where it's not "sold out". So we can assume that the market there is sat= urated now mostly. > > There remain those countries where Starlink isn't officially available ye= t. Some of these get roaming service, and I'm aware of at least one of thes= e where spectral capacity is uncomfortably near >(BTW: USAID was going to pay for a fibre cable there Where? >so China wouldn't, but I guess Elon doesn't want USAID to pay for the cabl= e so China can own it. While they're waiting for it, Starlink gets seen the= re as being unable to meet demand. It makes no sense to me. Incidentally, t= he country has one of the largest sovereign waters in the world and China i= s just waiting for the opportunity to get a naval base there I guess - on t= he far side of Guam. Bye America!). In general assuming elon cares about things at this level of grand strategy is probably a decent assumption and I kind of expect that once the slash and burn approach to things like USAID is exhausted that worthy projects will actually proceed faster. But anyway, you got a link to this project? > > Other countries where you can't get Starlink yet might contribute another= few million users - India in particular. But with Starlink not being able = to support particularly high user densities anywhere because of the spectra= l constraints, we're unlikely to see billions of customers there either. 30m is more than enough for a highly profitable entity. > On 6/02/2025 4:53 am, the keyboard of geoff goodfellow via Starlink wrote= : > > "He wants to take food off the table of people=E2=80=94hard-working peopl= e." > EXCERPT: > > Two new independent estimates of revenue from SpaceX's Starlink Internet = service suggest it is rapidly growing, having nearly tripled in just two ye= ars. > > An updated projection from the analysts at Quilty Space estimates that th= e service produced $7.8 billion in revenue in 2024, with about 60 percent o= f that coming from consumers who subscribe to the service. Similarly, the m= edia publication Payload estimated that Starlink generated $8.2 billion in = revenue last year. > > These estimates indicate that Starlink produced a few hundred million dol= lars in free cash flow for SpaceX in 2024. However, with revenues expected = to leap in 2025 to above $12 billion, Quilty Space estimates that free cash= flow will grow to about $2 billion. SpaceX is privately held, so its finan= cial numbers are not public. > > Growing subscribers > > By launching thousands of satellites and developing an Internet service b= ased in low-Earth orbit=E2=80=94where the proximity of satellites to the gr= ound provides significantly faster and lower latency service than satellite= s in geostationary space=E2=80=94SpaceX has already exceeded space-based co= mmunication networks developed earlier. > > At the end of last year, Starlink had 4.6 million subscribers. Quilty's d= irector of research Caleb Henry noted that the previous incumbent players, = Hughes and ViaSat, had a combined 2.2 million subscribers at their peak abo= ut half a decade ago, largely in North America, with some in South America = and a smattering in Europe. Starlink is expected to add another 3 million s= ubscribers this year alone. > > SpaceX has other significant lines of business, including government cust= omers, particularly the US Department of Defense, as well as maritime (75,0= 00 vessels equipped with Starlink as well as 300 cruise ships) and aviation= segments. > > "The key takeaway I want everybody to walk away with is, if SpaceX was bu= ilding the Starlink system to pay for a Mars colony, we've got evidence tha= t the company will generate the type of free cash flows from the business t= hat could pay for said endeavor," said Chris Quilty, co-chief executive and= president of Quilty Space... > > [...] > https://arstechnica.com/space/2025/02/starlink-profit-growing-rapidly-as-= it-faces-a-moment-of-promise-and-peril/ > > > -- > Geoff.Goodfellow@iconia.com > living as The Truth is True > > > _______________________________________________ > Starlink mailing list > Starlink@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink > > -- > **************************************************************** > Dr. Ulrich Speidel > > School of Computer Science > > Room 303S.594 (City Campus) > > The University of Auckland > u.speidel@auckland.ac.nz > http://www.cs.auckland.ac.nz/~ulrich/ > **************************************************************** > > > > _______________________________________________ > Starlink mailing list > Starlink@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink --=20 Dave T=C3=A4ht CSO, LibreQos