A decision made under fire

On the night of 13 June 2025, Israeli CIOs learned what their disaster-recovery documents were actually worth. Over the following twelve days Iran fired more than 550 ballistic missiles and over a thousand attack drones at Israel, striking the Haifa refinery complex and the Weizmann Institute of Science, where laboratories and years of irreplaceable research data burned (Wikipedia's account of the Twelve-Day War collects the strike record). Eight months later the war reached the cloud itself: in the renewed escalation of early 2026, Iranian drones damaged two Amazon Web Services facilities in the United Arab Emirates and one in Bahrain, knocking availability zones out of the me-central-1 and me-south-1 regions — the first publicly confirmed case of a US hyperscaler's data centres being put out of action by enemy fire, as Data Centre Dynamics reported in March 2026.

Against that backdrop, consider the decision facing the technology chief of a mid-sized Israeli firm — an insurer, a hospital group, a food manufacturer — in mid-2026. Option one: Amazon's il-central-1 region, opened in Tel Aviv in August 2023 with three availability zones and a promised $7.2 billion of investment through 2037. Option two: Google Cloud's me-west1 region, live since October 2022. Option three: a five-megawatt building in Petah Tikva run by a company most people outside Israel have never heard of, which will take the firm's entire VMware server room as-is, replicate it to a second site in Haifa and a third copy in Cyprus, answer the phone in Hebrew in seventeen seconds, and send a human being to the customer's basement if that is what the week requires.

Option three is Triple C Cloud Computing. The difference between what it charges and what the hyperscalers charge is not noise. It is one of the purest observable prices in infrastructure economics: the premium a customer will pay for jurisdiction, proximity and continuity in a country at war. This article prices that premium, traces where it comes from, and asks how much of it survives the decade.

One company on HaSivim Street, not a conglomerate

The identity question is worth settling first, because the record contains a small trap. Directory entries and older references sometimes gesture at a "Triple C / C.C.C. group", as though a holding structure sat behind the brand. The public evidence shows something simpler: one privately held company, טריפל סי מחשוב ענן בע"מ — Triple C Cloud Computing Ltd — registered as company number 511402547, a number whose vintage matches the firm's founding in 1989 (registry mirror). The founder, Rami Nachum, remains chief executive and owner nearly four decades later. The "C.C.C." is the domain name — ccc.co.il — not a conglomerate; according to the company's own lore, recorded on the Hebrew Wikipedia entry, the name abbreviates "Trip to the Peak". What looks from outside like a group is in fact one owner-managed firm of roughly a hundred people, by its own archived account, that has bolted new licences onto itself for thirty-five years: systems integrator in 1989, hosting operator, data-centre owner from 2008, licensed internet service provider from 2009, cloud provider, and briefly — disastrously — a television service between 2017 and 2018.

The network record confirms the single-firm picture. Triple C originates routes as AS50463 under the name TRIPLEC-ASN; RIPE's registry shows the holder as Triple C Cloud Computing Ltd., and routing views show roughly 79 IPv4 prefixes covering some 37,000 addresses, IPv6 alongside, transit from Telecom Italia Sparkle, Cogent and PCCW Global, a presence at the Israeli internet exchange, and — a small but telling housekeeping detail — valid RPKI origin attestations across the announced space. Individual prefixes carry working labels like "Triple C Network VMcloud" and "IBM Cloud Customers", the visible sediment of a business that hosts other people's platforms as well as its own. In PeeringDB the company lists a selective peering policy and discloses little else, which fits a wartime posture: the main website itself refused connections from outside Israel throughout the research for this article, reachable only through archived copies — a geo-fence that became standard practice across Israeli infrastructure after October 2023. Even Data Center Map now withholds precise coordinates for Israeli facilities, citing the security situation.

One episode from the corporate biography deserves more attention than it usually gets, because it reveals the owner's capital discipline. In 2017 Triple C launched a television service — channels plus video-on-demand at fifty shekels a month — into a market where every carrier was bleeding money on content to defend broadband share. It attracted roughly five thousand subscribers. In October 2018, barely a year in, Nachum shut it down. Contrast that with the incumbents, who carried loss-making TV units for a decade as strategic furniture. A company of a hundred people cannot afford strategic furniture, and the speed of that retreat is the best available evidence for how the same owner would handle a failing cloud line: quickly, and without sentiment. It also explains the shape of everything since — no adventures into content, mobile or overseas expansion, just adjacent licences bolted onto the same building, the same customers and the same support desk.

The physical estate, as documented before the shutters came down: a headquarters and data centre at 49 HaSivim Street in Petah Tikva's Kiryat Matlon industrial quarter, occupied since 2003, with dual power feeds totalling five megawatts, rack densities up to 12 kW, and construction described to the Tier IV and Tier II grades of the EIA/TIA-942-A standard; a disaster-recovery facility in Haifa; and a contracted third-party backup site in Cyprus (company profile). Note the geometry of that last item, because the rest of this article turns on it: Israel's flagship sovereign-cloud independent keeps its customers' last-resort copy outside Israel.

What the premium actually buys

Strip away the brochure language and Triple C sells four things.

First, jurisdiction. Data hosted on HaSivim Street sits under Israeli law and only Israeli law. No US CLOUD Act reach, no foreign court order served on a parent company in Seattle or Mountain View, no dependency on the goodwill of a firm whose employees have publicly campaigned against its Israeli government contracts. For a defence-supply-chain machine shop, a municipal utility or a law firm holding sensitive files, that is not an abstraction; it is the difference between a subpoena question and a non-question. Amendment 13 to Israel's Privacy Protection Law, in force since August 2025, sharpened the point by giving the Privacy Protection Authority real fining power for the first time, pushing thousands of mid-sized Israeli organisations to map, for the first time, exactly where their data lives and who can compel access to it.

Second, continuity engineering priced for a small country at war. The Petah Tikva production floor replicates to Haifa; the Cyprus copy answers the scenario in which both ends of the country are degraded at once. The service catalogue, archived in early 2024, reads like a continuity engineer's toolkit: disaster recovery as a service on Acronis, dedicated Synology backup racks inside the facility, SIM-backed internet lines that fail over to cellular when trenched fibre is cut, transmission services, a content delivery network, and a Radware web application firewall for the DDoS weather that accompanies every escalation.

Third, labour — specifically, Hebrew-speaking, locally accountable support labour, which is the least glamorous and most defensible part of the moat. Israel's Ministry of Communications measures telecom customer service, and Triple C has made a marketing identity out of the results: the regulator's surveys have repeatedly placed it first on answer times — 94% of calls answered within six minutes in one index round, average waits measured in seconds in later rounds — and at the bottom of the complaints table relative to size. The company's own timeline boasts "first in response, last in complaints" off the ministry's 2020 study. A hyperscaler's enterprise support plan costs 3–10% of monthly spend and escalates to a ticket queue in another time zone; Triple C's support desk is eleven kilometres from many of its customers.

Fourth, compatibility. Triple C was the first Israeli provider to earn VMware's Cloud Verified badge in 2020, and its core vCloud and pCloud products take a customer's existing VMware estate without re-architecture — the archived pCloud pitch is explicit: move the whole server room, keep paying monthly, the servers stay in Israel. For the thousands of Israeli mid-market firms that never intended to rebuild their ERP on cloud-native services, "lift it as it is and give me one bill in shekels" is the actual product. The hyperscalers' migration economics assume engineering time the customer does not have.

The switching costs run in an unusual direction, and the company has made them a selling point rather than a trap. On the consumer side Triple C built its ISP brand in 2009 on being the first provider with no lock-in and no exit fees — an inversion of the Israeli telecom norm of retention departments and cancellation ordeals. On the cloud side the stickiness is structural rather than contractual: once a firm's VMware estate, its backup routines, its DR drills and its SIM-backed failover lines all terminate in one Petah Tikva facility, leaving means re-testing a continuity architecture that took years to trust — a cost no discount from a rival easily covers. The asymmetry is deliberate. Cheap to join, expensive in confidence terms to leave, with the exit barrier made of the customer's own risk aversion rather than the vendor's lawyers. Hyperscalers achieve stickiness through egress fees and proprietary services; Triple C achieves it by being the counterparty a nervous organisation has already rehearsed emergencies with. In a country where the emergencies are real, that is the stronger glue.

Who pays? Not the government's big ministries — those went to Nimbus, as we shall see. Not the startups, who were born on AWS. The payers are the unfashionable middle of the Israeli economy: manufacturers, importers, clinics, insurers' subcontractors, municipalities' vendors, accounting and law practices, and a consumer flank of home-fibre subscribers in the tens of thousands. It is a clientele that values a phone number more than an API, and it is precisely the clientele the hyperscalers are worst at serving.

The arithmetic of a five-megawatt cloud

Triple C publishes no accounts, so the honest way to size it is to assemble the arithmetic from public parts and label each one. What follows mixes evidence and inference, and says which is which.

Start with the consumer internet business, where prices are public. Comparison sites list Triple C fibre at about NIS 115 a month for 500 megabits and NIS 75–95 for slower copper-era packages, in line with its rivals (SmartCut's listing shows the tiers). On the cost side, the wholesale price an ISP pays to ride Bezeq's fibre was NIS 72 per line per month until the Ministry of Communications cut it to NIS 58 in February 2026, fixed for five years. That is evidence. The inference: a fibre subscriber yields perhaps NIS 40–55 a month of gross margin before transit, support and marketing. The Hebrew Wikipedia entry credits Triple C with subscribers in the "tens of thousands"; take 40,000 as an illustrative midpoint and the consumer ISP generates very roughly NIS 50 million a year of revenue and perhaps NIS 15–25 million of gross margin — a real business, but a contribution engine, not the story. Its strategic function is different: the 2009 ISP licence and the retail base give Triple C end-to-end control of the wire into the customer's building, which is what makes the continuity products credible, and the February 2025 agreement to buy wholesale fibre access from Partner — Partner's own network passing 1.1 million homes plus its rights on Bezeq's 2.5 million — extends that reach without a shekel of civil works.

Now the cloud side. The facility ceiling is known: five megawatts across dual feeds, 12 kW rack maximums — call it 350–400 usable racks at claimed density, which at Israeli commercial colocation and managed-VM rates implies a theoretical revenue capacity in the low hundreds of millions of shekels, far above what the company plausibly bills. The one available topline estimate comes from a third-party data vendor, RocketReach, which pegs revenue at $27.5 million — around NIS 100 million — with 149 employees against the company's own "about 100". Treat both numbers as soft. But run the division anyway: $185,000–$275,000 of revenue per employee. That is the signature of a services-and-integration house with a cloud attached, not of a cloud platform with services attached. AWS generates several times that per head; so does any business whose margin lives in software rather than people. Triple C's margin lives in people — support seats answered in seventeen seconds, field engineers, VMware administrators — plus a paid-off building, plus resale margins on the Fortinet, Trend Micro, Radware, Synology, Acronis and Microsoft licences threaded through its catalogue (the Ingram Micro cloud-distribution awards on its own timeline tell you where much of the product actually comes from).

The support desk — the moat itself — can be costed too, and the exercise shows why competitors decline to copy it. Answering nearly every call within seconds, around the clock, for a customer base this size requires perhaps twenty to thirty staffed seats across shifts; at a fully loaded NIS 12,000–18,000 per seat per month, that is NIS 3.5–6 million a year of labour, or roughly 4–6% of the estimated topline, spent on a function the rest of the industry treats as a cost to be starved. That is inference built on standard Israeli support-labour rates, but the conclusion is robust to the inputs: the moat costs single-digit millions of shekels a year to operate and would cost a hyperscaler nothing to match technically — yet none does, because at hyperscale the arithmetic inverts, and seventeen-second Hebrew answering for 200-employee accounts can never clear a return threshold in Seattle. The most defensible asset in this business is a cost centre its larger rivals are structurally forbidden, by their own economics, from building.

The premium itself, then, is best expressed not per-vCPU — on raw compute Triple C cannot and does not try to beat a hyperscaler region whose owner enjoys 15–25% regional price dispersion and global scale — but as a bundle price: for a typical 30–60 VM Israeli mid-market estate, local managed cloud bids commonly land 20–40% above the raw infrastructure cost of the equivalent hyperscaler footprint, and roughly at parity once enterprise support, egress fees, migration engineering and a bilingual managed-service layer are added to the hyperscaler side. That range is inference from market behaviour, not a published tariff; the fact that the comparison must be made bundle-to-bundle is exactly why the premium has survived this long.

Nimbus moves the fence line

For most of Triple C's life, "the cloud is abroad" was its strongest sales line. Regulated and security-conscious Israeli organisations could not, or would not, put workloads in Frankfurt or Virginia, and the local providers collected the residency premium on everything those customers ran. Project Nimbus ended that world deliberately. In April 2021 Israel awarded Google and Amazon a NIS 4 billion (~$1.2 billion) framework to build in-country regions and carry the government itself — ministries, state companies, and the defence establishment — with contract terms that require data to remain inside Israel and bar the vendors from withdrawing service under boycott pressure. Google's me-west1 region opened in Tel Aviv in October 2022, projected by the company to add $7.6 billion to Israeli GDP by 2030; Amazon's il-central-1 followed in August 2023, with availability-zone sites reported around Shoham, Beit Shemesh and the Sharon.

The economic effect on local providers was surgical. Nimbus did not attack their prices; it dissolved their argument. Once the state itself certified that government workloads — including, per a contract reported by Time, Ministry of Defense consulting engagements — could run on American-owned clouds inside Israeli borders, "data residency" stopped meaning "local company". A bank's regulator, a hospital's counsel, an insurer's board could now satisfy the residency question with a hyperscaler invoice. Everything Triple C sells above bare residency — jurisdiction over the operator rather than the data-centre address, support labour, VMware continuity, wartime service — had to carry the whole premium on its own.

And yet the fence line matters in both directions. Nimbus explicitly keeps the most sensitive classified systems out of the public tender, inside defence-run facilities; and its terms concede the core of the sovereignty argument — Israel demanded contractual insulation from foreign pressure precisely because the operators are foreign. Every clause in that contract is a state-sized version of the question Triple C's salespeople ask a mid-market CIO: when politics or war turns hostile, who actually controls the switch? For customers too small to negotiate Nimbus-style terms — that is, every private mid-market firm in the country — the only way to buy those clauses is to buy a local operator.

War reprices the insurance

Between October 2023 and mid-2025, continuity migrated from the IT budget to the board agenda. The first weeks of the Gaza war brought mass reserve call-ups that hollowed out IT departments, sustained DDoS and wiper campaigns, and a quiet rush by Israeli firms to replicate data — some of it abroad, some of it into hardened local facilities. The underground operator MedOne, whose granite-tunnel sites host carriers and, reportedly, hyperscaler infrastructure, ran through the period with zero downtime and has since announced seven more sites; its marketing now leads with missiles, not megabits. The Twelve-Day War of June 2025 converted the remaining sceptics: when a ballistic missile can erase a research institute's archives, "our data is in Israel" is no longer automatically a comfort.

Then came the inversion. The 2026 round of the Iran war, in which drone strikes disabled AWS availability zones in the Gulf and Israel and the United States struck data centres in Tehran, taught the region two lessons at once. Lesson one: hyperscale regions are military targets now — Euronews quoted analysts calling data centres "the new target in modern warfare". Lesson two, subtler and more important for pricing: data-residency rules turned from shield into trap. As Tech Policy Press documented, Gulf customers legally confined to their national region could not fail over to healthy regions abroad when the local one was hit. Residency without dispersion is concentration risk wearing a compliance costume.

Read Triple C's architecture against those lessons and its odd geometry starts to look like the product's core rather than a footnote. Petah Tikva for production and jurisdiction; Haifa for domestic separation; Cyprus — forty minutes' flight, outside the missile envelope, inside a familiar legal orbit — for the copy that survives the unthinkable. The company is not selling "your data never leaves Israel". It is selling a graduated insurance policy: Israeli control of the operator, Israeli residency for the workloads that need it, and a foreign-soil escape copy for continuity. That is a more sophisticated product than the sovereignty slogan suggests, and after 2026 it is also a more honest one than a single in-country hyperscaler region, whose three availability zones share one national airspace. The premium did not disappear when AWS and Google landed; the war split it in two — a jurisdiction premium the hyperscalers partially neutralised, and a kinetic-continuity premium they arguably enlarged.

The supplier under the floor

The gravest threat to Triple C's economics in the record is not Amazon. It is Broadcom. The company's entire differentiated cloud — vCloud, pCloud, the Cloud Verified badge it advertises as a first-in-Israel distinction — runs on VMware, and since Broadcom's acquisition the terms of being a VMware-based cloud provider have been rewritten in ways that small operators worldwide have described as existential. Broadcom shut the VMware cloud-provider partner programme to all but an invited handful — industry accounts count the authorised list falling from roughly 4,500 providers globally to the low dozens by late 2025 — while licensing shifted to per-core subscriptions with steep minimums, and costs rose by multiples. For a provider whose pitch is "bring your VMware estate as-is", the licence bill is the cost of goods sold. Every renewal cycle now transfers some unknowable share of Triple C's premium from Petah Tikva to San Jose, and the alternatives — re-platforming its cloud onto Proxmox, OpenStack or Nutanix — would burn the very compatibility that anchors the offer.

The rest of the cost base is more conventional and more manageable. Transit comes from three global carriers plus the local exchange, a commodity in a country whose submarine-cable landings keep multiplying. Power — the binding constraint of the wider Israeli data-centre market — is comparatively trivial at Triple C's scale: five megawatts is a rounding error against the 313 megawatts a single developer has under construction, and a facility built in 2008 in an industrial park has long since amortised its shell. Wholesale last-mile access, the largest single line in the ISP business, just moved in the company's favour by regulatory fiat: the February 2026 tariff cut hands every small provider a NIS 14-per-line-per-month gift, explicitly framed by the ministry as protection for competition. Distribution risk (much of the security catalogue arrives via Ingram Micro) and Microsoft's cloud-solution-provider terms round out a supplier picture in which one dependency — VMware — dwarfs all others. A company that markets independence from foreign clouds is, under the floor tiles, deeply dependent on a single foreign software vendor in the middle of a historic price offensive.

A conviction in the customer file

The assignment any honest analyst must complete here concerns the public record of Triple C's conduct in exactly the market segment — institutional tenders — where its jurisdictional pitch should resonate most. In February 2017 Israel's competition authority indicted five computing firms and eleven executives for coordinating bids on public tenders worth some NIS 17 million between 2009 and 2012. In September 2019 the Jerusalem District Court convicted Triple C and its former sales vice-president of restrictive-arrangement and deceit offences for rigging four server-supply tenders run by Bezeq and the Israel Electric Corporation, with Rami Nachum convicted under officer-responsibility provisions; the sentence, handed down in 2020, was a NIS 200,000 corporate fine, NIS 100,000 plus a month of community service for Nachum, and seven months of community service for the former executive (Globes covered the indictment).

Three economic observations follow. First, the offence is a decade and a half old and pre-dates the cloud business; the court treated it at the misdemeanour end of cartel conduct, and nothing in the record since suggests recurrence. Second, it nonetheless functions as a standing tariff on the company's addressable market: Israeli public procurement scores integrity records, and a conviction — however dated — is a durable handicap in exactly the government and utility tenders where a sovereign-cloud independent would otherwise be strongest. It is probably not a coincidence that the public procurement record searched for this article shows no significant central-government cloud wins for Triple C, while the state's cloud money flowed to Nimbus. Third, and least comfortably for the sovereignty narrative: the case is a reminder that "local" is not a synonym for "trustworthy". The premium for local jurisdiction rests on the belief that a domestic operator is more accountable to the customer than a foreign one. The conviction shows the local market's own institutions doing that accountability work — which cuts both ways, validating Israeli enforcement while denting the halo of its beneficiary.

The wider evidence gap belongs in the same file: no audited financials, no disclosed customer list, no published uptime history, and defence-adjacent demand that can only be inferred from the product set (air-gapped-friendly private cloud, local clearance-holding staff, SIM-backed links) rather than named contracts. For a company whose product is trust under stress, the thinness of its public record is itself a pricing factor.

Squeezed from above, crowded from below

The competitive map has redrawn itself twice in five years. From above, the hyperscalers: two in-country regions, Nimbus branding, and the gravitational pull of every new SaaS and AI service landing there first. From below and beside, a construction boom without Israeli precedent: roughly 250 megawatts of grid-connected data-centre capacity today, with developers holding a further two gigawatts in planning — Mega Or's 313 megawatts under construction, Serverfarm and the Israel Infrastructure Fund's $1.5 billion, 130-to-200-megawatt campus at Ashdod, a $300 million Nebius AI facility at Modi'in, and a dozen newly arrived public companies chasing the same grid connections. CTech's survey of the boom carries investor warnings about solar-rush economics: capacity racing ahead of the grid, rents assumed rather than contracted.

Triple C competes with almost none of this directly — and that is the point. The boom is wholesale real estate and AI compute; Triple C is retail service. Its nearest true competitors are the other survivors of Israel's managed-hosting generation: MedOne with its underground tunnels and hyperscaler tenants, Bynet and the carrier-owned clouds of Bezeq International, Partner and Cellcom, all selling flavours of the same locality premium, several of them with deeper pockets and their own fibre. Against them Triple C's differentiators are its regulator-certified service record and its independence from any carrier's bundling agenda — an independence the ISP reforms have made double-edged. The "unified internet" reform let Bezeq sell infrastructure and service as one package, and analysts expected the incumbent to absorb half of its own ISP affiliate's customers within a year — a consolidation wave that squeezes every standalone ISP's consumer base even as the wholesale tariff cut fattens per-line margins. Triple C's answer, the Partner wholesale deal, effectively rents a second national fibre footprint from a rival rather than surrendering the retail relationship. It is the move of a company that understands its consumer arm is strategically, not financially, essential.

The substitute that matters most is none of these: it is the customer's own basement. Israeli mid-market firms kept server rooms long after their European peers, and every war scare pushes some back toward on-premises hardware plus a backup contract — a smaller, uglier version of the continuity product. Triple C's integration arm profitably serves even that choice, selling the servers, the Synology backups and the transmission line. The company is, in the end, indifferent to which flavour of "keep it close" the customer buys, so long as the customer keeps buying proximity.

Signals from the street

The unofficial record — forums, blogs, review aggregations, employee chatter — sketches a company whose service reputation is genuinely unusual for its sector and whose sharp edges are commercial rather than technical. Consumer forums have debated Triple C for two decades (one long-running FXP thread treats it as the connoisseur's ISP), and aggregated reviews on Mishtalem Li skew positive on responsiveness in a market where the default sentiment toward providers is contempt. The complaints that recur are about money, not packets: an HWzone thread warns fibre switchers about equipment charges, and a widely shared blog post recounts suing the company in small claims over NIS 32 while conceding its service was excellent — a perfect miniature of the pattern: superb phones, rigid billing. Employee reviews on TheWorker read like a family firm's: stability and informality praised, pay and hierarchy grumbled about.

What do these signals suggest, economically? That the service moat is real — corroborated independently by the regulator's own indices, which is rare; unofficial praise and official measurement pointing the same way is about as strong as evidence gets for a private company. That the billing rigidity is the sound of a low-margin retail arm defending pennies, consistent with the wholesale-tariff arithmetic above. And that the third-party revenue estimate of $27.5 million, while unverifiable, is at least not contradicted by headcount, facility scale or market position. What would settle the open questions: audited or registry-filed financials, a published customer-count for the cloud side, or a single disclosed reference contract from a regulated institution. None currently exists in public.

How long the premium survives

Assemble the pieces. Triple C's premium has three layers, and they erode on different clocks. The pure residency layer — data on Israeli soil — is already gone as a differentiator; Nimbus and two hyperscaler regions killed it between 2021 and 2023, and its remaining value accrues to anyone with an Israeli address, including Amazon. The jurisdiction layer — an operator that Israeli courts, and only Israeli courts, can compel — erodes more slowly, because no hyperscaler can offer it even in principle; its half-life is set by Israeli regulation, and the trend (Amendment 13, procurement sovereignty clauses, wartime information-control habits) is currently extending it, not shortening it. The continuity-and-labour layer — Haifa, Cyprus, seventeen-second phones, engineers who drive to the customer — is the most durable of all, because it is made of things hyperscalers structurally will not build for 200-employee customers, and after 2026 it is the layer the region's wars keep revaluing upward.

Against that stands the erosion working from inside: a supplier (Broadcom) capable of confiscating the VMware layer's economics at each renewal; a consumer ISP base exposed to incumbent bundling; an ownership structure — one founder-owner in his fourth decade, no visible succession — that concentrates both the firm's institutional memory and its key-person risk; and a scale ceiling, five megawatts in a market now measured in gigawatts, that caps the upside as surely as it limits the exposure. The likeliest trajectory is neither triumph nor extinction but a long, profitable narrowing: a NIS 100-million-class firm holding a defensible ledge — VMware-shaped, Hebrew-speaking, war-priced — while the mountain around it is rebuilt by companies a thousand times its size. Premiums like Triple C's rarely vanish; they retreat into the customers for whom no substitute exists, and get repriced upward on the way.

What would change the judgement

A handful of discoverable facts would move this assessment materially. Registry financials or an audited statement showing cloud and hosting revenue well above — or below — the $27.5 million third-party estimate would resize the whole franchise; the estimate is the softest number in this article. Evidence of a completed migration off VMware, or a disclosed post-Broadcom licensing arrangement, would defuse the largest single risk; conversely, a public price shock passed to customers would confirm it. A named win in a government, municipal or defence-adjacent tender would show the 2019 conviction's procurement shadow has lifted, and would reopen the market segment this analysis assumes is closed. Verified wartime performance data — uptime through June 2025 and the 2026 escalation, or a customer attestation of a Cyprus failover actually exercised — would convert the continuity premium from architecture into record. Succession or sale: any transaction involving the founder's holding would test whether the premium belongs to the company or to the man. And a strike, cyber or kinetic, on any Israeli civilian data centre would reprice every number in this piece overnight — in which direction depends on whose building is hit.

Sources and signals

The evidence behind this analysis is public, and the load-bearing items are listed here so the reader can weigh them directly.

Company identity and infrastructure:

Legal and regulatory record:

Market structure, Nimbus and the war:

Pricing, suppliers and street signals:

Where this article states a range or a multiple that appears in none of the above — subscriber midpoints, per-line margins, bundle-premium percentages — it is labelled as inference in the text, and the fastest way to falsify it is the financial disclosure the company has never made.