Thesis · Data Centers & Physical Infrastructure

Will CoreWeave's FY2026 Form 10-K retain a weighted-average or stated useful life of six years or more for its GPU computing equipment, with no AI-hardware impairment charge above $250M disclosed for FY2026?

Prior-generation accelerators keep recontracting at economic prices, so neocloud depreciation schedules hold and no material stranding charge appears.

6 dated tests · 0 resolved · next on

The tests More theses in Data Centers & Physical Infrastructure

Who is on each side

The two sides.

Each side in DeepStack’s words, then the voices who argue it, attributed and paraphrased, dated where the date is known. A voice links to its page on DeepStack where one exists.

The case for yes

GPU fleets keep recontracting, no stranding

Declining marginal performance gain per new generation plus supply scarcity keep A100/H100-class fleets commercially contracted past their assumed useful lives. CoreWeave disclosed an A100 contract running into 2029 on architecture from 2020, showing a prior generation recontracting rather than being written off. As GPU markets mature and the marginal performance improvement of new generations declines relative to cost, customers maintain existing hardware and upgrade at lower rates, so CoreWeave's FY2026 10-K retains a stated useful life of six years or more with no impairment charge above $250M.

The case for no

Obsolete GPU hardware gets stranded

As small local models improve in efficiency and capability faster than data center hardware depreciation schedules assume, an increasing share of inference workload migrates away from centralized data centers, undermining the revenue assumptions behind long useful lives. Equipment becomes obsolete in two to three years while data centers are built to operate for decades, creating a mismatch between financed life and actual economic life. Nvidia's model depends on making current hardware obsolete with faster successors, so CoreWeave's FY2026 10-K shortens its stated useful life below six years or discloses an AI-hardware impairment charge above $250M.

2 voices

GPU fleets keep recontracting, no stranding

  1. Ren

    CoreWeave disclosed an A100 contract running into 2029 on architecture from 2020. If hardware generations recontracts rather than being written off, and if this behaviour generalizes, then payback maths for neocloud fleets improve and the…

    Source

  2. a16z New Media

    As GPU markets mature and the marginal performance improvement of new generations declines relative to cost, customers maintain existing hardware and upgrade contracts at lower rates, sustaining a secondary market for older SKUs.

    Source

3 voices

Obsolete GPU hardware gets stranded

  1. Michael Burry

    As small local models improve in efficiency and capability faster than data center hardware depreciation schedules assume, an increasing share of inference workload migrates away from centralized data centers, undermining the revenue assumptions…

    Source

  2. Tscsw

    Physical infrastructure operates on time scales incompatible with both digital deployment speed and hardware obsolescence. Equipment becomes obsolete in 2-3 years but data centers operate for decades, creating a cascading problem where deployed…

    Source

  3. Vitaliy Katsenelson

    Nvidia's model depends on making current hardware obsolete with faster successors, and its sole-supplier position lets it charge software-like margins, so money spent now buys assets that lose value fast; data centers rushed up with unconstrained…

    Source

What would settle it

The dated tests.

Each test names the public source that settles it, the rule and the day. The Call’s number, the readers’ median and the machines’ numbers stand beside it, as on the Docket.

  1. Vertiv, purchases of property and equipment as a share of revenue, quarter ended September 30, 2026: whether it comes in at or above 4%.

    On the Docket Pendingth125-t01Pre-specified

    Question
    Did the purchases of property and equipment of Vertiv for the quarter ended September 30, 2026 come to at or above 4.0% of its revenue?
    Source
    SEC EDGAR, XBRL company facts of Vertiv (Forms 10-Q and 10-K)
    Rule
    Divide the purchases of property and equipment by the revenue that Vertiv reports for the quarter ended September 30, 2026 in its 10-Q or 10-K filing on SEC EDGAR. The answer is yes if the ratio is at or above 4.0%; no if it is below 4.0%.
    If the source is late, revised or silent
    If Vertiv has not filed by November 16, 2026, or a filing restates a period the rule reads, the test is inconclusive, and the Result says what was read and what was missing.
    Base rate
    Vertiv's purchases of property and equipment as a share of revenue was 5.3% for the quarter ended June 30, 2026, by its filing on SEC EDGAR.
    Outside view
    A year earlier, for the quarter ended June 30, 2025, the same line read 1.7%, so the rungs sit on the last print.
    What would change the number
    A change in product mix, a one-time charge, or a reclassified line would move the reading.

    The Call, before the dateExpects yes. Gives the yes answer 72%.

    72%

    0 · no100 · yes

    The machinesClaude 61% · DeepSeek 64%

    Claude: 61%DeepSeek: 64%
    The ladder 3 rungs on one reading, Oct 28, 2026: each asks whether it lands above its threshold, so the Call’s numbers down the rungs sketch where it expects the reading to land.

    1. Above 4%Call 72% · Claude 61% · DeepSeek 64%

      72%

    2. Above 5.2%Call 50% · Claude 60% · DeepSeek 7%

      50%

    3. Above 6.5%Call 27% · Claude 22% · DeepSeek 6%

      27%

    0 · no50100 · yes

    Your number

  2. Vertiv, purchases of property and equipment as a share of revenue, quarter ended September 30, 2026: whether it comes in at or above 5.2%.

    On the Docket Pendingth125-t02Pre-specified

    Question
    Did the purchases of property and equipment of Vertiv for the quarter ended September 30, 2026 come to at or above 5.2% of its revenue?
    Source
    SEC EDGAR, XBRL company facts of Vertiv (Forms 10-Q and 10-K)
    Rule
    Divide the purchases of property and equipment by the revenue that Vertiv reports for the quarter ended September 30, 2026 in its 10-Q or 10-K filing on SEC EDGAR. The answer is yes if the ratio is at or above 5.2%; no if it is below 5.2%.
    If the source is late, revised or silent
    If Vertiv has not filed by November 16, 2026, or a filing restates a period the rule reads, the test is inconclusive, and the Result says what was read and what was missing.
    Base rate
    Vertiv's purchases of property and equipment as a share of revenue was 5.3% for the quarter ended June 30, 2026, by its filing on SEC EDGAR.
    Outside view
    A year earlier, for the quarter ended June 30, 2025, the same line read 1.7%, so the rungs sit on the last print.
    What would change the number
    A change in product mix, a one-time charge, or a reclassified line would move the reading.

    The Call, before the dateEither answer fits the Call. Gives the yes answer 50%.

    50%

    0 · no100 · yes

    The machinesClaude 60% · DeepSeek 7%

    Claude: 60%DeepSeek: 7%

    Your number

  3. Vertiv, purchases of property and equipment as a share of revenue, quarter ended September 30, 2026: whether it comes in at or above 6.5%.

    On the Docket Pendingth125-t03Pre-specified

    Question
    Did the purchases of property and equipment of Vertiv for the quarter ended September 30, 2026 come to at or above 6.5% of its revenue?
    Source
    SEC EDGAR, XBRL company facts of Vertiv (Forms 10-Q and 10-K)
    Rule
    Divide the purchases of property and equipment by the revenue that Vertiv reports for the quarter ended September 30, 2026 in its 10-Q or 10-K filing on SEC EDGAR. The answer is yes if the ratio is at or above 6.5%; no if it is below 6.5%.
    If the source is late, revised or silent
    If Vertiv has not filed by November 16, 2026, or a filing restates a period the rule reads, the test is inconclusive, and the Result says what was read and what was missing.
    Base rate
    Vertiv's purchases of property and equipment as a share of revenue was 5.3% for the quarter ended June 30, 2026, by its filing on SEC EDGAR.
    Outside view
    A year earlier, for the quarter ended June 30, 2025, the same line read 1.7%, so the rungs sit on the last print.
    What would change the number
    A change in product mix, a one-time charge, or a reclassified line would move the reading.

    The Call, before the dateExpects no. Gives the yes answer 27%.

    27%

    0 · no100 · yes

    The machinesClaude 22% · DeepSeek 6%

    Claude: 22%DeepSeek: 6%

    Your number

  4. Super Micro Computer, long-term debt as a share of the quarter's revenue, quarter ended September 30, 2026: whether it comes in at or above 30%.

    On the Docket Pendingth125-t04Pre-specified

    Question
    Did the long-term debt of Super Micro Computer at the end of the quarter ended September 30, 2026 come to at or above 30.0% of its revenue for the quarter?
    Source
    SEC EDGAR, XBRL company facts of Super Micro Computer (Forms 10-Q and 10-K)
    Rule
    Divide the long-term debt that Super Micro Computer reports at the end of the quarter ended September 30, 2026 by the revenue it reports for the quarter ended September 30, 2026, in its 10-Q or 10-K filing on SEC EDGAR. The answer is yes if the ratio is at or above 30.0%; no if it is below 30.0%.
    If the source is late, revised or silent
    If Super Micro Computer has not filed by November 16, 2026, or a filing restates a period the rule reads, the test is inconclusive, and the Result says what was read and what was missing.
    Base rate
    Super Micro Computer's long-term debt as a share of the quarter's revenue was 36.5% at the end of the quarter ended June 30, 2026, by its filing on SEC EDGAR.
    Outside view
    A year earlier, at the end of the quarter ended June 30, 2025, the same line read 2.0%, so the rungs sit on the last print.
    What would change the number
    A change in product mix, a one-time charge, or a reclassified line would move the reading.

    The Call, before the dateExpects yes. Gives the yes answer 68%.

    68%

    0 · no100 · yes

    The machinesClaude 24% · DeepSeek 87%

    Claude: 24%DeepSeek: 87%
    The ladder 3 rungs on one reading, Nov 3, 2026: each asks whether it lands above its threshold, so the Call’s numbers down the rungs sketch where it expects the reading to land.

    1. Above 30%Call 68% · Claude 24% · DeepSeek 87%

      68%

    2. Above 40%Call 46% · Claude 22% · DeepSeek 6%

      46%

    3. Above 55%Call 23% · Claude 18% · DeepSeek 91%

      23%

    0 · no50100 · yes

    Your number

  5. Super Micro Computer, long-term debt as a share of the quarter's revenue, quarter ended September 30, 2026: whether it comes in at or above 40%.

    On the Docket Pendingth125-t05Pre-specified

    Question
    Did the long-term debt of Super Micro Computer at the end of the quarter ended September 30, 2026 come to at or above 40.0% of its revenue for the quarter?
    Source
    SEC EDGAR, XBRL company facts of Super Micro Computer (Forms 10-Q and 10-K)
    Rule
    Divide the long-term debt that Super Micro Computer reports at the end of the quarter ended September 30, 2026 by the revenue it reports for the quarter ended September 30, 2026, in its 10-Q or 10-K filing on SEC EDGAR. The answer is yes if the ratio is at or above 40.0%; no if it is below 40.0%.
    If the source is late, revised or silent
    If Super Micro Computer has not filed by November 16, 2026, or a filing restates a period the rule reads, the test is inconclusive, and the Result says what was read and what was missing.
    Base rate
    Super Micro Computer's long-term debt as a share of the quarter's revenue was 36.5% at the end of the quarter ended June 30, 2026, by its filing on SEC EDGAR.
    Outside view
    A year earlier, at the end of the quarter ended June 30, 2025, the same line read 2.0%, so the rungs sit on the last print.
    What would change the number
    A change in product mix, a one-time charge, or a reclassified line would move the reading.

    The Call, before the dateEither answer fits the Call. Gives the yes answer 46%.

    46%

    0 · no100 · yes

    The machinesClaude 22% · DeepSeek 6%

    Claude: 22%DeepSeek: 6%

    Your number

  6. Super Micro Computer, long-term debt as a share of the quarter's revenue, quarter ended September 30, 2026: whether it comes in at or above 55%.

    On the Docket Pendingth125-t06Pre-specified

    Question
    Did the long-term debt of Super Micro Computer at the end of the quarter ended September 30, 2026 come to at or above 55.0% of its revenue for the quarter?
    Source
    SEC EDGAR, XBRL company facts of Super Micro Computer (Forms 10-Q and 10-K)
    Rule
    Divide the long-term debt that Super Micro Computer reports at the end of the quarter ended September 30, 2026 by the revenue it reports for the quarter ended September 30, 2026, in its 10-Q or 10-K filing on SEC EDGAR. The answer is yes if the ratio is at or above 55.0%; no if it is below 55.0%.
    If the source is late, revised or silent
    If Super Micro Computer has not filed by November 16, 2026, or a filing restates a period the rule reads, the test is inconclusive, and the Result says what was read and what was missing.
    Base rate
    Super Micro Computer's long-term debt as a share of the quarter's revenue was 36.5% at the end of the quarter ended June 30, 2026, by its filing on SEC EDGAR.
    Outside view
    A year earlier, at the end of the quarter ended June 30, 2025, the same line read 2.0%, so the rungs sit on the last print.
    What would change the number
    A change in product mix, a one-time charge, or a reclassified line would move the reading.

    The Call, before the dateExpects no. Gives the yes answer 23%.

    23%

    0 · no100 · yes

    The machinesClaude 18% · DeepSeek 91%

    Claude: 18%DeepSeek: 91%

    Your number

In The Buildout

The stories that take up this thesis, with their own Calls and tests.

A thesis page states one question and both sides of it; DeepStack takes no side in words, and its view is the number it gives each test before the date. Analysis and opinion, not investment advice. Voices are paraphrased from public writing and attributed by name; DeepStack shows no score, ranking or accuracy for any person. Pre-specified means the words were approved on or before the test’s date. Specified after the date means they were written afterwards: such a test is counted and shown, and never rated. An open question has no words yet. The Call’s number is the chance, from 0 to 100, that the Call gives the yes answer to a test’s question, recorded before the test’s date; it is DeepStack’s own reading, not a market price and not advice. The readers’ mark on the same line is the readers’ median, the middle of the numbers signed-in readers gave the test, shown from 25 readers. The machines: Each model read this page and nothing else. The number that counts is the latest one each model gave before the test closed.