WYFI
WYFI · WhiteFiber, Inc. · Catalyst Calendar · as of August 18, 2026
Banded 0–3 / 3–6 / 6–12 months, earliest first. Numbering is gapless and runs straight through the bands, so a boundary never resets the count. An ID marks position in the current ordering and changes between revisions, so catalysts are named by title.
Service was expected to commence July 2026, subject to equipment delivery and acceptance milestones. Associated project-level financing was expected to close June 2026 from a binding term sheet. No confirmation of either was found as of August 3, 2026; the Q2 report is the likely disclosure venue.
Over $160M total contract value across five years, roughly $32M annually if spread evenly. The structure is attractive: an investment-grade named counterparty, twelve months of customer prepayments, third-party data center capacity so no owned capital expenditure, and project-level financing that limits reliance on the corporate balance sheet.
Commences on schedule with financing closed. First evidence that the prepayment-funded, project-financed model executes, and it diversifies revenue away from both the Initial Customer and Nscale at once.
Delivery or acceptance milestones slip, or the term sheet fails to convert. A second contract slipping in the same year as NC-1 shifts the read from one hard project to a pattern of optimistic timelines. Adds European regulatory and FX exposure.
Confidence is Medium because the company stated a month that has now arrived without public confirmation either way.
Source: Press release and Form 8-K, May 21, 2026. Filed
Twelve months after the Ordinary Shares began trading. Form S-3 generally requires at least twelve calendar months of Exchange Act reporting history and timely filings.
Eligibility does not create dilution; it removes friction. An issuer confined to Form S-1 raises equity slowly and visibly, while a shelf-eligible issuer can file an S-3 and an at-the-market program and issue opportunistically. Against cash and restricted cash of $60.4M at June 30, 2026 and quarterly capital expenditure of roughly $175M, the capacity matters more than it did when this card was written, and a working capital deficit of $28.6M is why. The full set of paths sits in the dilution overlay in section 5.
A shelf filed and left undrawn is optionality: it lowers perceived financing risk without diluting anyone. If the permanent NC-1 facility (S2) closes first, the shelf may be held as insurance rather than used.
A shelf filed and used at a depressed price. The stock is roughly half its late-June level, so issuing here is more dilutive than issuing at $46 would have been. Bit Digital's 70.1% stake also carries an eventual incentive to monetise, and a registered secondary is a distinct overhang from a primary raise.
Confidence is High on the eligibility date, which is mechanical. Whether a shelf is actually filed is unknown and unscheduled.
Source: IPO date per Form 8-K exhibit, August 8, 2025; eligibility criteria per SEC Form S-3 General Instruction I.A. Filed
Management stated on May 14, 2026 that it expected to begin delivering capacity during Q2 2026, with full revenue contribution beginning in Q3 2026. Original billing targets were 30 April and May 30, 2026. No public confirmation that billing has commenced was found.
Converts roughly $921M of remaining performance obligations from a promise into a recognized revenue stream. The backlog schedule implies about $50M of 2026 recognition and about $90M per year from 2027. This is the event the valuation is waiting on.
Both 20 MW phases energised close together, pulling 2026 recognition toward the disclosed $51.3M rather than a stub. Clean commissioning also validates the retrofit-speed claim underpinning the pipeline.
The medium-voltage switchgear issue persists or a new commissioning problem emerges. A slip into Q4 pushes meaningful revenue into 2027, extends the bridge-financing period at 9.5%, and invites questions about whether the six-month retrofit claim generalises from a 5 MW Quebec site to a 40 MW North Carolina one.
Confidence is Medium: the window is company-stated, but the original target has already been missed once by roughly a quarter.
Source: Form 8-K and press release, December 18, 2025; Q1 2026 press release and earnings call, May 14, 2026; Form 10-Q, May 14, 2026. Filed
Commercial launch targeted for Q3 2026. Full-fibre spectrum testing was stated as planned before launch. Announced July 9, 2026.
The cross-data-center networking architecture links two geographically separated facilities into a single logical GPU supercluster. Initial results: 111.2 Tbps across 83 km of dark fibre at 0.9 ms guaranteed round-trip latency, within 8% of the physical limit for light in fibre, using only part of the available spectrum. If commercialised, it is the company's only proprietary product.
Launch with a named customer, disclosed pricing, or a telecom or sovereign-AI partnership. The one item here that could re-rate the equity on something other than megawatts, giving it a technology argument rather than a property one.
Launch slips, or arrives with architecture detail and no customer. A capability announcement without commercial validation, from a company already under scrutiny for missed timelines, risks reading as narrative management. Full-spectrum testing has not been reported as complete.
Confidence is Medium: a company-stated quarter, recently reaffirmed, but for a first-of-kind product with a testing gate still open.
Source: Press release, July 9, 2026. Filed
Indirect but real. GTC events set sector sentiment and disclose GPU roadmap and availability. Vera Rubin was confirmed at GTC 2026 as entering full production with mass shipments from late 2026, directly relevant to a company whose cloud fleet is H200, B200 and GB200 and whose GPUs depreciate over five years. The European venue is also relevant to the Paris deployment (D1) and to sovereign-AI positioning.
Announcements reinforcing European sovereign-AI demand, or NVIDIA partnership news involving smaller infrastructure providers, tend to lift the whole neocloud complex. WhiteFiber has an existing NVIDIA relationship through its deployments.
An aggressive next-generation roadmap accelerates the obsolescence clock on installed H200 and B200 fleets, pressuring residual values and the five-year depreciation assumption, as set out in GPU supply and depreciation cadence (S5).
Confidence rates the event occurring, on a published date. No WhiteFiber-specific announcement is scheduled for it.
Source: NVIDIA published GTC schedule, retrieved July 28, 2026. Filed
Based on the November 13, 2025 precedent for Q3 2025. No company-confirmed date.
This should be the first quarter of full NC-1 revenue contribution, per management's framing on May 14, 2026. If Q2 shows a partial ramp, Q3 is where the run-rate becomes measurable and the roughly $90M annualised colocation figure can be checked against reported results rather than against the backlog schedule.
Colocation revenue at or above roughly $22–23M for the quarter annualises close to the backlog schedule and would be the first hard confirmation that the contract converts as disclosed. Positive operating cash flow driven by billings rather than prepayments would address the Q1 working-capital question.
A partial ramp well below the implied schedule, or evidence that pass-through electricity costs and NC-1 operating expense compress the contract's margin. Depreciation also steps up sharply once $294.3M of construction in progress is placed in service, and Q3 is likely the first quarter carrying its full weight.
Source: Prior filing dates per EDGAR; Q1 2026 press release, May 14, 2026. Filed
A company target of approximately 76 MW gross by Q4 2026, stated in the FY2025 10-K and reiterated in investor materials, measured against roughly 11 MW online at the end of 2025.
This is the second rung on the capacity ladder and the clearest single test of whether the retrofit model scales: a roughly sevenfold increase in energised capacity within twelve months, most of it NC-1.
Hitting it would place WhiteFiber among the faster converters of contracted capacity to live capacity in the peer group, where published activation rates have been reported in the 4–29% range. That comparison is the strongest available argument for a premium multiple.
Missing it. The target already carries a known drag: the MTL-2 build was put on hold to prioritize NC-1 and other projects, and further utility power allocations were described as subject to funding and permits. A miss would call the 1,500 MW pipeline into question further than it would call NC-1 into question.
Confidence is Low: a full-year capacity target dependent on the same commissioning process that has already slipped, with no interim milestone disclosure.
Source: Form 10-K, March 26, 2026; investor presentation furnished on Form 8-K, June 2026. Filed
San Jose. The first full post-Vera-Rubin GTC. It sets the GPU roadmap determining refresh economics for the cloud fleet and the density specifications colocation facilities must be engineered to. NC-1 is built for up to 150 kW per cabinet, high by current standards, and GTC is where whether that remains sufficient becomes visible.
Roadmap disclosures favoring high-density retrofit capacity, or continued emphasis on distributed and multi-site training that would validate the Project Redwood architecture (D4).
Density or cooling requirements advancing past what a retrofit envelope can economically accommodate, undermining the cost advantage the model rests on. Also a recurring sector-wide volatility event.
Source: NVIDIA published GTC schedule, retrieved July 28, 2026. Filed
Based on the March 26, 2026 precedent for FY2025.
The first audited full year containing NC-1, and the first with a genuine year-over-year comparison as a standalone public company, since FY2025 contained carve-out periods with allocated Bit Digital expenses. It is also where the Initial Customer concentration percentage gets restated.
Full-year revenue above the $79.2M FY2025 base with colocation the larger segment, positive adjusted EBITDA at scale, and concentration reduced. Audited confirmation of the contract economics would remove much of the current uncertainty discount.
Audit-driven surprises: goodwill or long-lived asset impairment on the $19.8M of goodwill or the GPU fleet, receivable write-downs following the DNA Fund pattern, or an internal-control finding. The internal-controls risk is not hypothetical: the four-month delay evidenced by the Form 4 batch (R8) is a documented process lapse, and Regulation S-K Item 405 disclosure of it is required here if not in the proxy.
Confidence is Medium: annual filing deadlines are statutory, but the specific date is extrapolated from one prior year.
Source: Form 10-K filing date, March 26, 2026; SEC periodic reporting deadlines for non-accelerated filers. Filed
Based on the May 14, 2026 precedent.
The cleanest available read on steady-state economics: a full quarter of NC-1 at contractual capacity, with the escalator structure and pass-through mechanics visible in reported margins rather than inferred from the contract description. Also the first quarter to lap the Q1 2026 comparison that included the Initial Customer pause.
Demonstrated colocation margins at scale, positive free cash flow, and a run-rate making the 2027 backlog schedule of about $90.7M look conservative rather than aspirational.
Full-quarter depreciation and interest expense on the completed asset base overwhelming the revenue step-up, leaving the company EBITDA-positive but persistently loss-making at the net line with no visible path to GAAP profitability.
Source: Prior filing dates per EDGAR. Filed
Structural items that shape the backdrop but do not trade in the window.
The December 18, 2025 announcement referenced planning for potential expansion toward roughly double the initial 40 MW deployment by the end of 2027. Nscale holds priority notification rights on future NC-1 capacity. No order has been announced, and any decision would need to land within this horizon to meet that timeline.
An order would roughly double the contracted backlog from a single counterparty, moving NC-1 toward its 99 MW Duke Energy capacity agreement, which targets up to 99 MW gross within four years of May 16, 2025. Management has separately indicated the site may support up to 200 MW over time.
The strongest possible signal about the first 40 MW, since it would mean the customer is satisfied enough to double down. It would use already-secured power and an already-owned building, so incremental economics are far better than a new site.
Expansion deepens rather than diversifies concentration: the same counterparty, the same site, the same undisclosed end customer. It would also require fresh capital. And Nscale declining to expand, after holding priority rights, would be a quiet but pointed verdict on Phase 1 performance.
Confidence is Low: an aspiration referenced in a press release, contingent on NC-1 initial revenue confirmation (D3) succeeding and on financing that does not yet exist.
Source: Press release, December 18, 2025; Form 10-K, March 26, 2026, for the Duke Energy capacity agreement terms. Filed
Ongoing and undated, most material first. These take "Why undated" in place of timing confidence. Four of them were carried as dated items in the source numbering; an item with no date is a standing condition however it was previously counted, and the concordance in section 8 maps the change.
Roughly $921M of $923.7M in remaining performance obligations rests on one agreement with Nscale, a privately held company. WhiteFiber describes the arrangement as backed by an investment-grade hyperscaler offtake, but the end customer is not named in the disclosures reviewed. Separately, the Initial Customer represented about 70.7% of 2025 cloud-heavy revenue and has paused services.
The impact is total. Both segments carry near-total single-customer dependency at once: colocation on Nscale, cloud on the Initial Customer. The credit chain runs through a party investors cannot examine to another party they are not told the identity of.
Nscale is a well-regarded neocloud named among sector leaders and has raised substantial capital. If the undisclosed offtake is genuinely investment grade, the contract's credit quality may be better than the disclosure permits investors to verify. A closed permanent financing (S2) would be indirect third-party confirmation.
Any renegotiation, delay or default anywhere in the chain changes the backlog from an anchor into a question, and investors would likely learn of it after the fact. The DNA Fund termination, which left about $7.3M of receivables of which about $2.1M had been collected as of the 10-K, is a live precedent for how counterparty failure presents here.
Source: Form 10-Q, May 14, 2026; Form 10-K, March 26, 2026; Form 8-K, December 18, 2025. Filed
Originally expected to be formalised in early Q1 2026 per the December 18, 2025 announcement. As of May 27, 2026 the company described itself as still advancing non-dilutive permanent financing solutions.
Closing releases the Bit Digital security over Enovum NC-1 Topco equity, ends the 1.1x minimum-return bridge economics, and is the precondition for committing capital to a second site. The spread at which it prices is itself the signal: an independent read on the Nscale contract's credit quality.
Closes at a tight spread to yield on cost, as management originally guided, with an accordion feature. That validates the contract's bankability, cuts interest expense, and removes the governance discomfort of borrowing from the controlling shareholder.
Continued delay, or closing at a wide spread with restrictive covenants. Lenders declining to underwrite against the Nscale contract at reasonable terms would be an independent verdict on the counterparty, more informative than management commentary. Prolonged reliance on the parent facility compounds at 9.5% plus a 3% original issue discount.
This downside became more likely in late July 2026. AI-infrastructure credit repriced visibly: CoreWeave was reported on 29 July as seeking to raise the rate on a $2.6bn loan tied to Anthropic capacity, its 2032 bond yielding around 10.32% at a B rating, with technology credit default swap costs widening. WhiteFiber is placing project debt into a market actively repricing this risk, as set out in neocloud sentiment and credit conditions (S4).
Source: Press release, December 18, 2025; press release and Form 8-K, May 27, 2026; Bit Digital Form 8-K, May 20, 2026 Filed; sector credit reporting, July 29, 2026 Market
WYFI trades as part of a correlated complex, alongside CoreWeave, Nebius, IREN, TeraWulf, Cipher, Applied Digital and adjacent names, whose multiples move together on sector news regardless of company-specific developments.
The effect is demonstrated and large. On July 1, 2026, reporting that Meta was exploring sales of surplus GPU capacity under a Meta Compute initiative removed roughly 13–15% of the market value of CoreWeave and Nebius in one session. WYFI fell from roughly $46 in late June to $23.70 on 31 July with no company-specific announcement identified. Peer drawdowns from peak as of late July: CoreWeave about negative 61%, IREN about negative 47%, Nebius about negative 37%.
Continued hyperscaler capital expenditure and constrained power supply keep contracted capacity scarce. Sector re-ratings lift small caps disproportionately, and WYFI has shown a 4x range within twelve months. It has at points held up better than larger peers, plausibly because its backlog is contracted colocation rather than GPU rental.
Competition has broadened: SpaceX has entered alongside Meta, reportedly taking capacity deals from Google, Reflection AI and Anthropic. Two well-capitalized entrants in one year shift the question from share-taking to what happens when the largest infrastructure owners sell capacity directly; they build for peak training demand, sit at 30–50% utilization between runs, and carry none of the GPU-collateralised debt smaller operators rely on. Credit has also stopped being a forecast: the 2026–2028 refinancing wall is visible in prices, and although the convertible runs to 2031 and is not exposed to that wall, the price of capital for the category is set in this market and the NC-1 facility (S2) must be placed into it.
Source: Sector and credit-market reporting, Q2 to Q3 2026, including July 29, 2026; Synergy Research neocloud market forecast. Market
The cloud segment owns and leases NVIDIA hardware, in H200, B200 and GB200 deployments, depreciated over five years against a generation cycle that is faster than five years. The company also carries a variable GPU server lease costing $3.7M in Q1 2026, and disclosed $67.9M of deposits for equipment not yet delivered.
This determines whether reported earnings are economically meaningful. If useful life is shorter than the depreciation schedule, current losses understate true economic cost; if residual values hold, they overstate it. The Q1 sale of 126 H200 GPUs for $26.1M against a $24.3M carrying value is a single favorable data point rather than a pattern.
Sustained GPU scarcity supports residual values and rental pricing. Selling above book demonstrates the fleet is a liquid asset that can be recycled into capital, which is useful for a company this cash-constrained.
Vera Rubin mass shipments from late 2026 compress demand for prior-generation hardware. A residual-value reset would hit the balance sheet and the cloud economics at once, and cloud is still roughly three quarters of revenue. Supply risk cuts both ways: the medium-voltage switchgear issue that delayed NC-1 shows equipment availability affecting the colocation side as well.
Source: Form 10-Q, May 14, 2026, Notes 6 and 7; Form 10-K, March 26, 2026; NVIDIA GTC 2026 disclosures. Filed
At the July 29, 2026 meeting the FOMC held the target range at 3.50–3.75% for a fifth consecutive meeting, on a 9–3 vote with three members dissenting in favor of a hike. Markets now price two 25bp increases in 2026. CME FedWatch put a 41.9% probability on a hold in September, up from 24% the prior day. The driver is energy-led inflation associated with the US–Iran conflict.
Remaining dated events: the Jackson Hole symposium on 27–29 August, with Chair Warsh speaking and closely watched given his stated preference for less forward guidance; the FOMC on 15–16 September, described as the first meaningful test; then 27–28 October and 8–9 December. September and December carry a Summary of Economic Projections.
Two channels, both pointing the same way. Directly, WhiteFiber is a capital-intensive borrower that must close project financing (S2) and may access equity markets (D2), and rates set the price of both. Indirectly, unprofitable long-duration growth equities are among the most rate-sensitive assets in the market, and this company's value rests on contracted cashflows that mostly begin in 2027.
The conflict de-escalates, energy prices normalize, and the hold extends through year-end as some forecasters expect. That removes a headwind rather than creating a tailwind: the case for cuts has largely disappeared from market pricing.
A September or October hike raises the cost of the permanent NC-1 facility precisely while it is being negotiated, compounds the credit repricing described in neocloud sentiment and credit conditions (S4), and compresses the multiple on 2027 revenue that does not yet exist. Rate risk and sector credit risk are not independent: a hike would tighten AI-infrastructure lending further, hitting the financing (S2) through two channels at once.
Jackson Hole and the three remaining meetings are consolidated into this one condition rather than taking separate catalyst numbers. They are the same exposure, and separate entries would double-count it.
Source: Federal Reserve FOMC statement and implementation note, July 29, 2026, and published FOMC calendar Filed; CME FedWatch as reported July 30, 2026 Market
Operations span the United States (North Carolina, under a Duke Energy capacity agreement targeting up to 99 MW gross within four years of May 16, 2025), Canada (Quebec), Iceland, France, Japan and Australia. The company's risk disclosures flag power price and supply volatility, export restrictions and tariffs particularly with Canada, AI-specific regulation, and national security regulation across multiple jurisdictions.
Power availability is the binding constraint in this industry and is granted by utilities and regulators on their own timetables. Additional utility power allocations were described as subject to funding and permits. The company is also exploring natural gas fuel cell generation at certain sites to add power outside the utility queue.
Already-secured power is the scarce asset. Duke Energy has delivered 54 gross MW with a 99 MW agreement in place, and management believes the site may support up to 200 MW over time. Multi-jurisdiction presence also positions the company for sovereign-AI demand, which the Paris contract (D1) begins to demonstrate.
Data center power allocation has become politically contested in several jurisdictions, including Quebec where much of the Enovum estate sits. Tariffs on Canadian equipment and cross-border operations are a named company risk. FX translation produced a $(1.97)M other-comprehensive loss in Q1 2026 alone. US electricity prices rose roughly 6.3% in the twelve months to January 2026, more than double headline inflation at the time; colocation passes electricity through, so NC-1 and MTL-3 are largely insulated, but cloud is not, and cloud remains roughly three quarters of revenue.
Published estimates of current crude prices conflict, some placing Brent below $75 and others above $100, so this input is genuinely uncertain rather than settled. The same energy prices feed the inflation driving the rate path (S6).
Source: Form 10-Q forward-looking risk summary and MD&A, May 14, 2026; Form 10-K, March 26, 2026. Filed
Management stated in Q3 2025 that it expected to formalise its next development site in response to specific customer demand, and in May 2026 described advancing multiple larger site opportunities. The item is gated by the permanent financing (S2).
The pipeline is the third rung on the capacity ladder and carries no contracts, financing or power agreements. A single site converting from under review to acquired with an anchor tenant is what would make it a real asset rather than a slide.
A site announced with power secured and an anchor customer attached, in the NC-1 pattern. Management has been explicit that it wants demand-led site selection rather than speculative land banking, which is the more disciplined approach.
A site acquired speculatively before NC-1 is proven and financed, consuming capital the balance sheet does not comfortably have. Equally, prolonged silence would suggest the pipeline is not converting, and the pipeline is doing real work in supporting the current multiple.
Source: Q3 2025 results commentary, November 13, 2025; Q1 2026 press release, May 14, 2026; Form 10-K, March 26, 2026. Filed
WhiteFiber is a Cayman Islands company that listed in August 2025, and no notice of an annual general meeting or proxy filing was found as of July 28, 2026. For reference, majority owner Bit Digital announced its own 2026 annual general meeting on June 17, 2026.
Modest in isolation, but it is the only governance event on this calendar and the natural venue for any share authorisation increase, equity plan expansion beyond the current 4,000,000-share pool, or board changes. With Bit Digital holding 70.1%, outcomes are effectively predetermined, which is itself the point worth noting.
Addition of genuinely independent directors, or disclosure of a related-party transaction review process covering the Bit Digital facility. Either would partly address the governance discount a controlled company carries.
A proposal to increase authorised shares or expand the equity plan, read as preparation for dilution. Reliance on Nasdaq controlled-company exemptions, if disclosed, would confirm reduced independence requirements. The proxy is also where Regulation S-K Item 405 requires disclosure of delinquent Exchange Act Section 16(a) filings, and the two Form 4s filed roughly four months late (R8) should appear there.
Source: EDGAR filing history for CIK 0002042022, reviewed July 28, 2026; Bit Digital press release, June 17, 2026. Filed
Closed items, kept for the record, ordered by resolution date. A resolved catalyst takes the next free R number; it does not carry its old number over. Where an item spans dates the resolution date is the date it completed. An item whose completion date is not stated sorts last and says so.
9,375,000 Ordinary Shares priced at $17.00, trading from 7 August and closing August 8, 2025 for $159.4M gross. The full 1,406,250-share over-allotment was exercised on September 2, 2025, bringing total gross proceeds to approximately $183M. B. Riley Securities and Needham were joint bookrunners; Macquarie Capital also book-ran.
This establishes the reference price. The stock traded above $40 in October 2025 and below $11 at its low. Full exercise of the over-allotment indicates genuine demand at pricing.
Source: Form 8-K exhibit 99.2, August 8, 2025; Form 10-Q, May 14, 2026, Note 1.
Enovum NC-1 BIDCO executed a master services agreement with Nscale Services US Inc. and Nscale Global Holdings. The initial service order covers 40 MW IT load in two 20 MW phases, approximately $865M over an initial ten-year term including 3% annual escalators and non-recurring installation charges, excluding pass-through electricity and property taxes, on a modified gross lease.
This single contract is essentially the entire $923.7M backlog and the reason the market capitalization sits where it does. It is signed and disclosed; the open question is delivery, carried as NC-1 initial revenue confirmation (D3), not existence.
Source: Form 8-K and exhibit 99.1, December 18, 2025; Form 10-K, March 26, 2026.
Private placement of $230.0M of 4.500% convertible senior notes due 2031, priced January 26, 2026 with an initial conversion price of $25.91, a 27.5% premium at pricing. Net proceeds were approximately $222.1M. Roughly $120.0M was used for a zero-strike call option over 5,905,511 Ordinary Shares, lifting the effective conversion price to approximately $37.
This funded the NC-1 build without immediate equity dilution, at a coupon far below the related-party bridge that followed. It also consumed $120M of the proceeds on the call structure, additional paid-in capital fell from $504.7M to $388.9M as a result, and it is why the balance sheet now carries $222.3M of convertible notes net of issuance costs.
Source: Form 10-Q, May 14, 2026, Notes 2 and 12; FY2025 press release, March 26, 2026.
The Saint-Jérôme, Quebec facility was leased from April 11, 2025 under a 20-year term with a fixed-price purchase option. WhiteFiber became reasonably certain of exercising in December 2025, remeasured the arrangement as a finance lease, notified the lessor on December 31, 2025, exercised on January 14, 2026, and completed the purchase on May 8, 2026 at a price disclosed at approximately $53M.
This converts a leased revenue-generating asset into an owned one with expansion optionality. It is also the site demonstrating the six-month retrofit claim, and the one now supporting Cerebras. The item sorts on its 8 May completion rather than its 14 January exercise.
Source: Form 10-Q, May 14, 2026, Note 6; Q1 2026 press release, May 14, 2026.
A five-year agreement to provide AI compute infrastructure for an investment-grade technology customer in the Paris region using NVIDIA GPU systems, with total contract value in excess of $160M. Third-party data center capacity in France was secured and a binding term sheet was signed for project-level financing.
This is the first material contract that is neither Nscale nor the Initial Customer, and the only diversification event on the record. The structure is the template management says it wants to repeat: prepayments plus project financing, minimal corporate balance sheet. Delivery is carried as the Paris deployment go-live (D1).
Source: Press release and Form 8-K, May 21, 2026.
Enovum NC-1 Venture LLC entered a delayed-draw term loan facility of up to $100M, expandable to $150M, with Bit Digital Capital. It is guaranteed by White Fiber Operating Partnership LP and secured by equity in Enovum NC-1 Topco. B. Riley Securities acquired a $20M portion by assignment on identical economic terms with a 90-day term. Obligations and liens release upon permanent project financing.
This cuts two ways. It secured liquidity when the permanent facility had not closed, which is a real achievement. It also confirmed that the permanent facility had not closed, at expensive affiliate-negotiated terms. Full terms sit under Bit Digital control and related-party dependency (S3).
Source: Press release and Form 8-K, May 27, 2026; Bit Digital Form 8-K, May 20, 2026.
Initial testing of a proprietary distributed GPU supercluster architecture achieved 111.2 Tbps across 83 km of dark fibre, described as roughly double the capacity of comparable published full-spectrum field trials while using only a portion of available spectrum, at 0.9 ms guaranteed round-trip latency.
This resolved as an announcement rather than as an executed transaction. These are internal R&D results with no third-party verification, no customer and no revenue. The commercial event is Project Redwood commercial launch (D4), not this.
Source: Press release, July 9, 2026.
Four Form 4s were accepted by EDGAR on July 28, 2026, all reporting performance-based RSU vestings under the 2025 Omnibus Equity Incentive Plan. Confirmed: CEO Samir Tabar, 41,982 RSUs, transaction dated March 31, 2026, valued at $11.91, holdings to 222,087; CFO and director Erke Huang, 41,982 RSUs, same 31 March date and price, holdings to 196,311; Tabar again, 7,079 RSUs, transaction dated July 24, 2026, valued at $38.85, the June 30, 2026 closing price, holdings to 229,166. A fourth filing was submitted in the same batch and its contents were not individually verified.
Two things follow. A performance milestone was met on June 30, 2026: the 7,079-unit grant vested on achievement of an unspecified milestone, valued at the 30 June close. The milestone is not identified, so it cannot be tied to any other item here, but a compensation committee certified that something was achieved as of the quarter end. Separately, Exchange Act Section 16(a) requires a Form 4 within two business days, and two of these report 31 March transactions filed on 28 July. These were non-discretionary plan vestings rather than open-market trades, so this is a controls matter rather than a trading-conduct one, and it must be disclosed under Regulation S-K Item 405 in the annual report or proxy. It contrasts with an otherwise clean periodic record, since no NT 10-K or NT 10-Q has been filed.
Source: Forms 4 accepted by SEC EDGAR July 28, 2026, 17:25 to 17:26 ET, CIK 0002042022; Bit Digital Form 4 batch accepted the same evening at 18:59 ET.
Duke Energy completed the work required to deliver the initial 54 gross MW of utility power to the Madison, North Carolina site, supporting the planned 40 MW IT load. Generators, UPS systems and chillers were all reported on site, with approximately 600 personnel working through final commissioning.
This removes the hardest risk from the NC-1 story. Grid interconnection is the binding industry constraint and typically the item that slips by years. What remains is commissioning, a real problem but a categorically smaller one.
The source records the date the delivery was confirmed rather than the date the work completed, so this item sorts last within the resolved class rather than at 14 May. Establishing the completion date is not established.
Source: Q1 2026 press release and earnings call, May 14, 2026.
The question was renegotiation or permanent loss, and the answer is permanent loss. The Company executed a Termination Agreement with the Initial Customer during the second quarter of 2026 and redeployed the GPUs previously allocated to it. This was the customer that represented approximately 70.7% of 2025 revenue, contracted for 2,048 GPUs over three years at an estimated $50M of annualised revenue.
The settlement is better than the headline loss suggests and is not yet cash. The agreement preserved $12.5M of previously invoiced unpaid trade receivables, and that balance was fully collected by June 30, 2026. A $2.2M bad debt expense was recognised on the unpreserved remainder. Separately the customer owes a fixed termination fee of $12.3M, recognised as revenue in the quarter, which was amended after quarter-end to $15.7M and remains outstanding as of the date of the Form 10-Q.
Recognising a termination fee as revenue is the single most consequential accounting fact in this quarter, and it is why the Q2 2026 results (R11) read the way they do. It is not recurring, it is not yet collected, and it is 42.6% of reported revenue.
Residual risk: the amended $15.7M is a receivable from a counterparty that has just walked away from a three-year contract, and no payment date is disclosed. The replacement of the revenue is partly under way rather than done: new customer orders contracted in the six months and through the filing date total $635.8M of contracted revenue over six-month to three-year terms, which is the number to watch against this loss.
Source: Form 10-Q for the quarter ended June 30, 2026, filed August 12, 2026, customer agreements and revenue notes. Filed
Filed on the third day of the ~9 to 13 August window this card named, with no separate results release: the Form 10-Q is the disclosure. Consensus was roughly negative $0.41 per share on about $19.4M of revenue.
| Line | Q2 2026 | Q1 2026 | Reading |
|---|---|---|---|
| Revenue, as reported | $28.8M | $21.9M | Beats the $19.4M consensus by 49% |
| Less the termination fee | $12.3M | n/a | One-off, recognised as revenue, and not yet collected |
| Revenue, excluding it | $16.5M | $21.9M | Below consensus, and down 24.6% sequentially |
| Cloud, excluding it | $11.5M | $16.8M | Down 31.5%, a second consecutive quarterly fall |
| Colocation | $4.7M | $4.8M | Flat. The step-up this card looked for did not arrive |
| Loss per share | ($0.39) | n/a | Against consensus of $(0.41), and flattered by the same fee |
The card named nine conditions in advance and they are scored against what it asked, not against what the quarter contained. No upside limb landed cleanly: NC-1 billing was not confirmed, the colocation increase being attributed to MTL-3 rather than to Nscale; colocation did not step up; the Initial Customer did not return but terminated (R10); receivables were collected, accounts receivable falling from $91.7M at March 31 to $23.2M and six-month operating cash flow running at positive $89.1M, against a $2.2M write-off on the departed customer. Three of five downside limbs landed: cloud declined a second consecutive quarter, NC-1 slipped again, and draws on the Bit Digital facility got heavier, at $20M on July 27 and $10M on July 31. Two did not: general and administrative expense fell to $14.8M from the $17.8M run-rate the card named, and receivables converted rather than aged.
The balance sheet is where the quarter is hardest. Working capital swings from $85.2M at December 31, 2025 to a deficit of $28.6M at June 30, 2026. Cash and restricted cash fall to $60.4M, of which $56.1M is unrestricted, from $80.1M at March 31. A $5.0M impairment of capitalised software is taken in the quarter. Against that, Adjusted EBITDA is positive at $5.5M against $3.3M a year earlier, and the loss of $15.0M is struck after $6.6M of depreciation and the impairment.
Residual risk: the fee will not recur, so the Q3 comparison starts from $16.5M rather than $28.8M unless the new contracts bill in time. A working capital deficit alongside a $15.7M receivable from a departed customer is the specific thing to watch, and the equity issuance route it points at is why the Form S-3 shelf eligibility card (D2) matters more now than when it was written.
Source: Form 10-Q for the quarter ended June 30, 2026, filed August 12, 2026. Filed Consensus per Benzinga and Unusual Whales, retrieved July 28, 2026. Market
Every category reviewed appears as a row, including those that produced catalysts, which read across by title and ID. A category checked against a named source and found empty reads None. A category never examined reads Not established, which is a gap rather than an absence.
| Category | Status | Basis |
|---|---|---|
| Earnings | Covered above | Q2 2026 results (R11), Q3 2026 results (D6), FY2026 results (D9), Q1 2027 results (D10). |
| Product launches | Covered above | Project Redwood commercial launch (D4). |
| Major contracts | Covered above | Nscale expansion order (D11); Paris deployment go-live (D1) delivers the Paris AI compute agreement (R5). |
| Macro events | Covered above | Federal Reserve policy path (S6), carrying the September, October and December FOMC dates and the Jackson Hole symposium. |
| Industry conferences | Covered above | NVIDIA GTC Berlin (D5), NVIDIA GTC 2027 (D8). |
| Regulatory decisions | None | None pending: no agency approval, license decision or enforcement matter identified. Diffuse regulatory exposure is carried at power availability, energy prices, tariffs and cross-border regulation (S7) instead. |
| Lawsuits | None | Law-firm investigation notices and the Legal Proceedings item of the Q1 2026 10-Q were checked; no filed securities class action, derivative action or announced investigation was located. |
| Management changes | Covered above | Announced July 30, 2026 and filed August 3, 2026. Erke Huang resigned as chief financial officer, principal financial and accounting officer and director, effective 1 August, becoming senior adviser and non-voting board observer while remaining chief financial officer of Bit Digital; Justin Zhu succeeded him at $450,000 a year, having resigned his own posts at the parent. Director agreements were also filed for Samir Tabar and Michael Rulf, an amendment for Pruitt Hall, and a professional services agreement with Rasterwerk LLC, a company of Mr Rulf's. Officers are now Tabar (CEO), Zhu (CFO), Billy Krassakopoulos (President) and Thomas Sanfilippo (CTO). Tabar is also CEO of Bit Digital, carried at Bit Digital control and related-party dependency (S3). |
| Investor days | None | None scheduled or announced. An investor presentation was furnished on Form 8-K in June 2026, which is not a scheduled event. |
| Share buybacks | None | No program announced or authorised. The company is a net consumer of capital. |
| Dividends | None | None declared and no dividend policy disclosed. |
| Mergers and acquisitions | None | None pending: no transaction is outstanding. The MTL-3 purchase (R4) was an option exercise on an existing lease, already completed. |
| Index inclusion | Not established | Not examined. Cayman incorporation and a 70.1% controlled float complicate eligibility for several major indices, but no analysis was reviewed. Also named in what was not checked. |
| Lock-up expiry | None | No forward catalyst arises. A customary 180-day IPO lock-up from August 7, 2025 would have expired around February 2026. |
Sits across every other catalyst rather than beside them. 38,848,118 Ordinary Shares were outstanding as of August 10, 2026, from 38,841,201 at June 30 and 38,614,216 at May 12, 2026. The paths below are drawn from filings; none is a prediction that issuance will occur.
| Instrument or path | Potential shares | Trigger | Status |
|---|---|---|---|
| 2031 convertible notes, $230M at $25.91 | ~8.88M | Conversion; the zero-strike call lifts the effective price to approximately $37 | Outstanding. Stock at $25.95, four cents above the conversion price |
| Zero-strike call option | 5,905,511 | Equity-classified hedge, purchased for approximately $120M | In place; reduces without eliminating net dilution |
| 2025 Omnibus Equity Incentive Plan | 4,000,000 | RSU and option grants; the CFO received 66,094 RSUs on March 19, 2026 | Active; vesting schedules run through at least October 2028 |
| Form S-3 shelf and ATM | n/a | Eligibility from approximately August 7, 2026, carried as Form S-3 shelf eligibility (D2) | No shelf filed as of August 3, 2026 |
| Bit Digital secondary sell-down | ~27.0M | Registered resale or block trade | No announcement; the 70.1% stake is a standing overhang |
| Bit Digital term loan expansion | n/a | Expandable from $100M to $150M by mutual agreement | Debt rather than equity: no share-count dilution, but 1.1x minimum return plus 9.5% is a drag on equity value |
The 8.88M potential conversion shares are roughly 23% of the current count.
The key asymmetry is that the convertible sits above the current price, so it is not near-term dilutive, which also means it currently functions as straight debt with a 2031 maturity rather than as equity in waiting. The dilution that matters most is the one not yet on the table: a shelf drawn at a depressed price. Because the stock has roughly halved since late June, the number of shares required to raise any given sum has roughly doubled. Dilution risk here is inversely related to how well the near-term operational catalysts resolve.
What would prove this read wrong, stated in advance. Each trigger is settleable from a document; an unquantified threshold is not a test. Each test is built from a claim this file already makes.
| # | If this happens… | …this was wrong | Status |
|---|---|---|---|
| 1 | The Q3 2026 Form 10-Q reports colocation revenue below $22M for the quarter. | That NC-1 converts at the run-rate the backlog schedule implies. This file states that $22–23M for the quarter would annualise close to that schedule. | Untriggered |
| 2 | The FY2026 Form 10-K reports revenue recognized against the colocation backlog below $51.3M for 2026. | That the disclosed remaining-performance-obligation schedule is a reliable guide to timing. The file takes $51.3M as the 2026 figure. | Untriggered |
| 3 | The company discloses energised capacity below 76 MW gross at December 31, 2026. | The capacity ladder as stated, and with it the claim that the retrofit model scales at the rate management has targeted. | Untriggered |
| 4 | WYFI closes above $25.91. | That the convertible is not near-term dilutive and currently functions as straight debt rather than equity in waiting. | Untriggered |
| 5 | An NT 10-K or NT 10-Q is filed on EDGAR under CIK 0002042022. | That the periodic-filing record is clean and the reporting lapse is confined to Exchange Act Section 16(a). | Untriggered |
| 6 | A Q3 2026 Form 10-Q reports general and administrative expense above $17.8M. The Q2 limb is settled: $14.8M, tested and not triggered. | That the Q1 2026 general and administrative step was a one-off public-company build rather than a new run-rate. | Untriggered |
| 7 | A Form 8-K reports the NC-1 permanent facility closing at a coupon at or above 9.5%. | That permanent financing replaces the related-party bridge at a better cost. The bridge carries 9.5%. | Untriggered |
One further test was attempted and could not be built to a settleable threshold. It is recorded as a failure in section 8 rather than reworded into something that only looks settleable.
Every ID in one table, gapless within each class. Links point at title slugs so they survive renumbering.
These items are not independent shots on goal. Three chains connect most of them, and reading the calendar as a flat list overstates the number of genuinely separate paths.
The financing loop. The NC-1 permanent project financing (S2) releases the Bit Digital delayed-draw facility (R6), which frees capital for the next development site (S8), which supports the Nscale expansion order (D11). Every downstream item is gated by one overdue event. The permanent facility was expected in early Q1 2026 and had not closed as of the last disclosure found. Until it does, the company pays bridge pricing on its flagship asset and has less room to commit to a second site.
The reflexive loop. NC-1 initial revenue confirmation (D3) feeds the quarterly results (R11 and D6), which feed the share price, which sets the pricing available under Form S-3 shelf eligibility (D2) and the convertible economics, which sets the cost of the capital funding the 76 MW gross capacity target (D7). This runs both ways. A clean NC-1 start makes equity cheaper to raise and pushes the stock further above the $25.91 conversion price, converting debt to equity on the company's terms; a miss makes any raise more dilutive precisely when more capital is needed. The stock fell from roughly $46 in late June to $23.70 on 31 July, including a 9.71% drop that Friday from a $27.06 open, so the loop currently points the wrong way. With a float of about 11.1 million shares against 38.6 million outstanding, moves in both directions are amplified.
The offset problem. The Initial Customer renegotiation (R10) determines whether the colocation revenue confirmed at NC-1 (D3) is additive growth or replacement growth. This is the chain most easily missed. The FY2025 10-K disclosed that the Initial Customer represented approximately 70.7% of 2025 revenue and had paused services pending renegotiation, with GPUs redeployed to other clients and no definitive resolution. If that customer does not return on comparable terms, the roughly $90M-per-year colocation ramp implied by the backlog schedule backfills a hole in cloud rather than stacking on top of it, and consolidated revenue growth would then understate the operational achievement, or overstate it, depending on which side was expected.
One consequence is worth stating plainly. The Q2 2026 results (R11), NC-1 initial revenue confirmation (D3), the Initial Customer renegotiation (R10) and the NC-1 permanent project financing (S2) all resolve, or fail to resolve, within roughly the same six-week window around the Q2 report. That is unusual concentration, and it means the August to September period carries far more of this calendar's total variance than its share of the timeline suggests. Two of those four carry no date at all.
Maps the numbering used before this revision to the current IDs, so earlier log entries remain resolvable. Four items previously carried as dated catalysts are undated by substance and are now standing conditions.
| Was | Now | Catalyst |
|---|---|---|
| 1 | R11 | Q2 2026 results and Form 10-Q |
| 2 | D3 | NC-1 initial revenue confirmation |
| 3 | R10 | Initial Customer renegotiation resolution |
| 4 | S2 | NC-1 permanent project financing close |
| 5 | D1 | Paris deployment go-live and project financing close |
| 6 | D4 | Project Redwood commercial launch |
| 7 | D2 | Form S-3 shelf eligibility |
| 8 | D5 | NVIDIA GTC Berlin |
| 9 | D6 | Q3 2026 results and Form 10-Q |
| 10 | D7 | 76 MW gross capacity target |
| 11 | S8 | Next development site from the 1,500 MW pipeline |
| 12 | S9 | First annual general meeting and proxy statement |
| 13 | D9 | FY2026 results and Form 10-K |
| 14 | D8 | NVIDIA GTC 2027 |
| 15 | D10 | Q1 2027 results |
| 16 | D11 | Nscale expansion order toward double the deployment |
| S1 | S4 | Neocloud sentiment, hyperscaler self-supply and AI credit conditions |
| R10 | S1 | The concentration chain: Nscale and the undisclosed end customer |
| S2 | S6 | Federal Reserve policy path |
| S3 | S3 | Bit Digital control and related-party dependency |
| S4 | S5 | GPU supply, obsolescence and depreciation cadence |
| S5 | S7 | Power availability, energy prices, tariffs and cross-border regulation |
| C1 | R1 | Initial public offering |
| C2 | R2 | Nscale master colocation agreement |
| C3 | R3 | $230M convertible notes and zero-strike call |
| C4 | R4 | MTL-3 purchase option exercised and closed |
| C5 | R9 | Duke Energy delivers initial 54 gross MW to NC-1 |
| C6 | R5 | Paris AI compute agreement signed |
| C7 | R6 | $100M delayed-draw facility from Bit Digital Capital |
| C8 | R7 | Project Redwood R&D results announced |
| C9 | R8 | Four Form 4 insider filings, two materially delinquent |
| Level | Means |
|---|---|
| High | Date is company-announced, protocol-defined, or statutorily fixed. |
| Medium | Date inferred from filing cadence or a stated deadline window. |
| Low | Date is a judgment call. Could move by a quarter or more. |
| n/a | A standing condition with no diary date. Takes "Why undated" instead. |
| Tag | Means |
|---|---|
| Filed | Stated in an SEC filing or company press release, cited by form and date. Also covers a dated official publication by a named non-SEC issuer, such as the NVIDIA GTC schedule or a Federal Reserve statement, with that issuer named in the Source line. It does not cover a third party's characterisation of another body's facts, however authoritative that body: that is Market. |
| Estimate | Derived or inferred here rather than reported. |
| Open | Expected or targeted but unconfirmed, with nothing filed either way. |
| Market | Price, float, consensus and peer market data, stamped with the close or report date. |
Defects the document has not resolved, with the text at issue quoted and the reason each remains open.
A falsification test on customer concentration failed and is not published. The only formulation this file offers is, in the FY2026 results upside: "Full-year revenue materially above the $79.2M FY2025 base with colocation the larger segment, positive adjusted EBITDA at scale, and concentration meaningfully reduced." No threshold is stated for what counts as reduced concentration, and the file states no target percentage anywhere. None is derivable from the sources this document carries. The governing requirement is that every trigger must name an observable event settleable from a document, and that a threshold which cannot be settled is not a test; rewording the phrase into something that merely looks settleable would defeat the purpose of writing the test in advance.
Impact levels are a construction, not a source figure. The source recorded impact as prose without a High, Medium or Low scale. The levels shown on each card and in the summary rank the file's own impact statements, for example "the single highest-impact event on this calendar" against "modest on its own". No impact assessment was added beyond ranking what the document already asserts, and the ranking is recorded here so it is visible as a judgment rather than read as a source figure.
A dependency reference in the source pointed at the wrong item. The financing chain referred to the 9.5% related-party bridge by a bare number that, under the source's own numbering, belonged to an unrelated conference catalyst. The chain now names each item by title with its current ID in parentheses, which resolves the reference without changing the claim. This is the failure mode that naming by title exists to prevent.
Confidence rates timing, never outcome. A High-confidence catalyst can be a coin flip; a Low-confidence one can be near-certain in direction.
Impact is independent of confidence. A High-impact, Low-confidence item is the most important kind here: it matters enormously and could land any time. The 76 MW gross capacity target (D7) and the Nscale expansion order (D11) are both of that kind.
Primary filings and company releases first, with form type and date.
SEC EDGAR filings for WhiteFiber, Inc., CIK 0002042022: Form 10-Q for the quarter ended March 31, 2026 (filed May 14, 2026); Form 10-K for the year ended December 31, 2025 (filed March 26, 2026); Form 10-Q for the quarter ended September 30, 2025; Forms 8-K dated August 8, 2025, December 18, 2025, May 21, 2026, May 27, 2026 and June 10, 2026; four Forms 4 accepted July 28, 2026; Form 3 filings, August 2025; Form S-1 registration 333-288650 and the related final prospectus. Bit Digital, Inc., CIK 0001710350: Form 8-K dated May 20, 2026 and Form 4 batch dated July 28, 2026. Company press releases dated September 17, 2025, November 13, 2025, December 18, 2025, March 26, 2026, May 14, 2026, May 21, 2026, May 27, 2026 and July 9, 2026. Q1 2026 earnings call transcript, May 14, 2026.
Secondary: Federal Reserve FOMC statement and implementation note, July 29, 2026, and published FOMC calendar; NVIDIA published GTC schedule; Benzinga, Unusual Whales, Stocktwits and ChartMill for earnings-date estimates; market data as of the August 3, 2026 close; CME FedWatch as reported July 30, 2026; Synergy Research and trade press for sector and credit-market context.
Reviewed August 1, 2026 against CIK 0002042022, recorded so that nothing found is distinguishable from not checked.
| Form | Status | Most recent, and note |
|---|---|---|
| 8-K | Found | Jun 10, 2026, investor presentation. |
| 10-Q | Found | May 14, 2026, Q1 2026. |
| 10-K | Found | Mar 26, 2026, FY2025. |
| Form 4 | Found | Four filed Jul 28, 2026, two delinquent by roughly four months, carried at the Form 4 insider filings (R8). |
| Form 3 | Found | Filed at IPO, Aug 2025. None since, consistent with no new officers or directors. |
| S-1 | Found | Registration 333-288650, Jul 2025. None since. |
| 424 | Found | Final IPO prospectus, Aug 2025. None since, so no follow-on offering. |
| S-3 | None | No shelf on file. Eligibility begins approximately Aug 7, 2026, carried at Form S-3 shelf eligibility (D2). |
| 13D / 13G | None | One beneficial-ownership filing exists and it is now read: a Schedule 13D filed by Bit Digital, Inc. on August 8, 2025 reporting 27,043,750 ordinary shares, approximately 74.3% of the shares outstanding and of total voting power at the offering, with sole voting and dispositive power. No amendment and no other filer appears across the complete index of 73 filings, searched August 3, 2026. The earlier statement that none was located was wrong: the filing is captioned Schedule 13D in full rather than abbreviated, and a search for the abbreviated caption alone does not return it. |
| NT 10-K | None | FY2025 annual report filed without a late notification. |
| NT 10-Q | None | Q1 2026 quarterly report filed without a late notification. |
| 6-K, 20-F, NT 20-F | Not applicable | Foreign private issuer forms. Despite Cayman Islands incorporation the company reports as a US domestic filer on 10-K, 10-Q and 8-K, so these will not appear. |
| 40-F, NT 40-F | Not applicable | Canadian multijurisdictional disclosure system forms. Not applicable to this filer. |
The periodic-report record is clean and the reporting lapse is confined to Exchange Act Section 16(a) beneficial-ownership reporting.
Index inclusion was not examined. Cayman incorporation and a 70.1% controlled float complicate eligibility for several major indices, but no analysis was reviewed and none is inferred here. Its row in section 4 reads Not established for that reason.
No earnings date on this calendar is company-confirmed. The quarterly and annual items are extrapolated from prior filing dates or taken from third-party estimators that disagree with each other: Benzinga gives August 9, 2026 for Q2, Unusual Whales 13 August, Stocktwits 14 August and ChartMill September. Four providers, four answers, so the windows carry the weight rather than the dates.
The completion date of the Duke Energy power delivery (R9) is not stated in any filing reviewed, which records only the date it was confirmed. That item therefore sorts last within the resolved class, and the actual completion date is not established.
The contents of the fourth Form 4 in the 28 July batch were not individually verified. Balance-sheet data derives from the March 31, 2026 10-Q and is four months old; the company was consuming cash rapidly and has facilities available, so the current cash and debt position is likely different. Peer credit data, including the CoreWeave bond yield, its rating and the reported loan repricing, comes from market reporting rather than from primary filings reviewed here. Published estimates of current crude prices conflict.
The Initial Customer is unnamed in the disclosures reviewed, as is the hyperscaler behind the Nscale offtake, so two of the three most important counterparties here cannot be independently assessed. Items marked unconfirmed, namely NC-1 billing commencement, the Paris go-live and the June financing close, may have occurred without the disclosure being found; absence of confirmation is not evidence of failure.
Analyst coverage is conflicted in at least one case. B. Riley Securities was an IPO bookrunner, publishes research on the company, and holds a $20M assignment of the related-party term loan. Published targets range from $20 to $50, a spread wide enough that consensus carries little information. Market
Third-party filing aggregators lag, and one produced an error recorded in the document log below. Aggregator fields reporting a most recent filing are not a substitute for checking EDGAR by form type, and Form 3, Form 4, 13D, 13G, 424 and NT filings are where they lag most.
Newest first. The original build entry is never removed or rewritten.
This document cited two Forms 8-K that were never filed. The sources list named Forms 8-K dated March 26, 2026 and May 14, 2026. Those are the dates WhiteFiber filed its Form 10-K and its Form 10-Q, both cited correctly elsewhere in the same list, and both also the dates of a company press release. No Form 8-K was filed on either date. The two dates are removed from the Form 8-K list; the five that remain are as filed, and no other source in the list changes.
The dilution table said the stock was below the conversion price, and at this document's own basis it is above. The 2031 notes convert at $25.91 and the pricing basis is the August 3, 2026 close of $25.95, so the cell reading "Stock at $23.70, below the conversion price" carried the 31 July close through the reprice of 3 August. It now reads $25.95, four cents above. The gating chain said a clean NC-1 start pushes the stock toward that conversion price and now says further above it. The economic reading is unchanged and is not restated: the zero-strike call lifts the effective price to about $37, which this document says in two places.
The sources list dated this document's market data to the 31 July close. The masthead has priced off the August 3, 2026 close of $25.95 since that date, and the capitalization of $1,008M is struck on it. The line now names August 3. No catalyst resolved, none was added and no expected date moved.
Three mentions of the 31 July close are left alone, being the fall from roughly $46 in late June to $23.70 on 31 July, described twice, and the 9.71% drop that Friday. Each dates itself and each is correct as history.
Not absorbed, and named instead. A Form 8-K of August 17, 2026 and the company's announcement the same day of an agreement to acquire two North Carolina development sites were not read. The submissions index was re-read on August 18, 2026 and the issuer newsroom was reached.
Q2 2026 was filed on August 12, 2026, inside the window this calendar named. Revenue $28.8M against a consensus near $19.4M, loss $0.39 per share against $(0.41). Carried as Q2 2026 results and Form 10-Q (R11).
The beat is a termination fee. $12.3M of that revenue is a one-off fee from the Initial Customer, recognised as revenue in the quarter. Excluding it, revenue is $16.5M, below the consensus it appears to have beaten and down 24.6% sequentially, and cloud revenue falls from $16.8M to $11.5M. The downside limb this calendar wrote in advance, cloud declining a second consecutive quarter, landed and is invisible in the headline.
The Initial Customer question is answered, and the answer is termination rather than renegotiation. Carried as Initial Customer agreement terminated (R10), which leaves the standing conditions and takes the earlier resolved number because it resolved inside the quarter. The fee was amended after quarter-end to $15.7M and remains outstanding; $12.5M of preserved receivables was collected and $2.2M was written off.
Working capital swings from $85.2M to a deficit of $28.6M between December 31, 2025 and June 30, 2026, with cash and restricted cash at $60.4M against $80.1M at March 31. A $5.0M software impairment is taken. Adjusted EBITDA is positive at $5.5M.
Four events after the quarter are recorded that no earlier revision could have carried. On August 10 the Company bought 55,105 SAIHEAT Limited shares for ~$1.0M at $18.15 and surrendered its $1.0M Canopy SAFE the same day; four Bit Digital managers, including its Chief Financial Officer, invested $0.5M each in that same PIPE personally. On July 30 a subsidiary leased data centre space in Sydney, a new country. On July 6 Enovum entered a syndicated RBC facility, with the prior agreement repaid in full on July 15 and CAD 36.8M drawn. On July 27 and 31 the Company drew $20M and $10M on the Bit Digital Capital facility.
The Form S-3 shelf eligibility date has passed and no shelf has been filed. That card moves from pending to live: eligibility is now a fact rather than a forecast, and against a working capital deficit it reads differently than it did.
The Central Index Key is recorded for the first time. Filings are swept against 0002042022 rather than against the ticker, which does not survive a rename. The index holds 78 filings to August 12, 2026, with no overflow file.
No rotation away from the colors that carry meaning here was needed.
Nothing else changed. No catalyst was added, removed or re-dated. The pricing basis is unchanged at the 3 August close, no session having completed since.
The masthead market capitalization was tagged as published market data and is computed in this document. It read Market and now reads Estimate: the figure is the 31 July close multiplied by the filed share count, derived here rather than published by a data provider.
The beneficial-ownership row was wrong, not merely unverified. It read that none had been located and that the position was unverified rather than confirmed absent. Bit Digital filed a Schedule 13D on August 8, 2025: 27,043,750 ordinary shares, sole voting and dispositive power, approximately 74.3% of the shares outstanding and of total voting power at the offering. The filing is captioned in full rather than abbreviated, which is why a search for the abbreviated caption returned nothing. The row now reads None, on the strength of a complete index of 73 filings searched today, and the holding is recorded.
This entry’s own composition line was wrong. It read 16 dated, 6 standing against the 12 and 10 that the cards, the summary table and the masthead all hold: the four items reclassified as undated by substance moved, and the line did not follow them. It now reads 12 dated, 10 standing, 9 resolved. No catalyst, figure or date changes; the count of items was misstated, not the items.
No catalyst changed state. Every falsification trigger was re-read limb by limb: three join two conditions with or, and in each case both halves name a filing, a threshold or a price, so none rests on a limb that could be argued either way.
The mark has changed since the note was written, so its description is updated, but it still yields no usable hue.
Repriced to the August 3, 2026 close, and the management-change row is no longer a null. The 3 August session has closed, so the basis moves July 31, 2026 → August 3, 2026 and the price $23.70 → $25.95, up 9.49% on the day. The filing record was swept again on 3 August by issuer central index key against the complete index of 73 filings. The newest item is the Form 8-K of that date, and this document now records what it contains rather than pointing at the companion: the null row for management changes read None, none has been announced, which the same day's filing had already made false. The horizon band boundaries do not move, being a convention this document adopted rather than a measurement.
Revised against the 28 July build. Repriced to the 31 July close of $23.70 (down 9.71% on the day, from a $27.06 open); market cap approximately $915M; float of approximately 11.1M shares added.
Filings. Form-by-form EDGAR audit across 8-K, 6-K, 10-Q, 10-K, 20-F, 40-F, S-1, S-3, 424, 13D, NT 10-K, NT 10-Q, NT 20-F, NT 40-F, Form 3 and Form 4; results tabled in Section 7. Correction: the 28 July build reported no filings since the 10 June 8-K, relying on a third-party aggregator that was stale. Four Form 4s were filed on July 28, 2026, two of them delinquent by roughly four months. Added C9; completed-event count 8 to 9. The periodic-report position (8-K, 10-Q, 10-K) was correct as originally stated and is unchanged.
Cards changed. S3 reframed: the 29 July FOMC held 9 to 3 with three dissents favoring a hike and market pricing moved from cuts to increases, reversing this card's original direction; Jackson Hole 27 to 29 August folded in. S1 updated for SpaceX's entry alongside Meta and the visible repricing of AI-infrastructure credit. S6 retitled and extended to electricity-price inflation. S4 updated for the shared compliance function with the parent. #4 downside expanded for the credit shift. #12 updated for Item 405 delinquency disclosure, #13 for internal controls. Summary-table skews for S1 and S3 moved to Downside, with a note on their correlation.
Unchanged. Items 2, 4 and 5 remain unconfirmed; no new 8-K, 10-Q or press release since 10 June and 9 July respectively.
Calendar built. 16 dated catalysts, 6 structural conditions, 8 completed events, 9 null categories. Priced off the July 27, 2026 close of $23.15.