← All thesesCBRSUnattractiveMedium convictionFriday 13 November 2026עברית

Cerebras Systems

This is the archived full research. Read the story

A maker of one giant AI chip, cut from a whole silicon wafer, that answers questions from AI models faster than anyone and rents that speed, mostly to OpenAI.

Our viewUnattractive Medium conviction
The scores
The profile · 1-4
Growth3/4 ↑

Revenue rose 76% in 2025 and 74% in Q2 2026, but two related buyers paid 86% of 2025 and OpenAI now carries the backlog.

Business quality3/4 ↓

Measured fastest per user and chosen by OpenAI, but NVIDIA bought Groq's design and reportedly serves OpenAI's newest fast tier on GPUs.

Financial position3/4 →

About $7.7B of net cash and an undrawn revolver, but the 750 MW build needs capital and leases the filings do not total.

Execution record3/4 →

Beat its core guide in all three quarters as a public company and raised the year; three quarters is too short a record for more.

Per share capture2/4 →

OpenAI's warrant, Amazon's and the founders' units could add about 14.7%, and warrant discounts cut revenue to 2031.

Potential3/4 →

A $26.1B take-or-pay backlog is committed and dated, a third due within two years; the inference market is large, its fast part uncounted.

Risk profile2/4 ↓

One customer holds most of the backlog, two related Abu Dhabi buyers most past revenue, and a short seller disputes the chips' life.

Valuation1/4 ↑

About 24.1 times sales, 308% of a three-stage value near $32.83 at an 8% return: priced for perfection, beyond the 20% margin.

Evidence quality3/4 →

The contract, backlog and margins are filed, but the threats rest on reported claims neither company confirmed.

The bet · 2-3 years
Safe, little upside1.0×

If it works 1.0× · if it fails you keep about 92%.

The Verdict

We rate Cerebras Unattractive, with medium conviction. Its chip is the fastest at answering AI questions, and the third quarter beat the company's own guide, raised the year and added Microsoft, but OpenAI still fills most of a $26.1B order backlog, NVIDIA is closing the speed gap and the losses run deep, so the business is mixed. On our base path the price is about 308% of our value; only if the backlog lands on its filed schedule does our value pass the price, hence medium conviction. As a bet on the big cloud providers it is defensive: the OpenAI business and the cash hold most of the value.

The long caseIf buyers keep paying more for faster answers, the clouds that serve them will want the fastest chip, and Cerebras could rent through every big cloud rather than one lab.

How we weighed it

The price against our value

Against it

Price at 308% of our value

The price sits far above our value, below only the case where the backlog lands on schedule.

The business

Cuts both ways

Mixed

A mixed business: the fastest chip, but one buyer.

Growth against what the price assumes

In its favour

The backlog grew and a third of it is now due within two years.

What could go wrong

Against it

One buyer carries it, NVIDIA is closing the gap and the chips may wear out early.

Against competitors

Cuts both ways

No filed rival compares directly; it has traded only briefly.

How its group is behaving

Cuts both ways

AI chip names move together on spending news.

The latest news and report

Cuts both ways

The report beat the guide and signed Microsoft, but the call said OpenAI's first tranches are behind schedule, with delay credits already taken off the backlog.

The evidence under this - 7 claims
C07facthigh confidence
In December 2025 OpenAI committed to buy 750 MW of Cerebras inference capacity on take-or-pay terms, deployed in tranches from 2026 to 2028 on three- or four-year terms extendable to five, with an option for 1.25 GW more by 2030.
against this
  • Revenue from OpenAI was $56.8M in Q2 2026, small beside the commitment, so delivery is still ahead (S02).
  • SemiAnalysis said OpenAI's newest fast tier runs on NVIDIA GPUs (C22).
  • On the Q3 call Cerebras said it is behind schedule on the first tranches, and either side may cancel a tranche delayed 90 days or more (C49).
  1. 2026-11-13editedre-read against the Q3 call: the finance chief restated the 750 MW take-or-pay terms and the 1.25 GW option; the delivery delay is recorded in C49
current state · 2026-11-12written 2026-11-03
C06facthigh confidence
Remaining performance obligations were $26.1B at 30 September 2026, mostly OpenAI's; Cerebras expects about 33% to become revenue in the 24 months to September 2028 and 44% in months 25 to 48, against 22% and 43% at 30 June.
against this
  • Delivery timing may change at the customer's request, the filing says.
  • The figure excludes most future pass-through costs, which are billed on top.
  1. 2026-11-12editedre-read in the Q3 10-Q: the backlog grew and more of it is scheduled inside two years
current state · 2026-09-30written 2026-11-03
C10facthigh confidence
MBZUAI and G42, related parties in Abu Dhabi, were 62% and 24% of FY2025 revenue; in Q3 2026 Customer A, related to G42, was 38% and an unnamed Customer B 34%, and two customers held 80% of receivables at 30 September.
  1. 2026-11-12editedre-read in the Q3 10-Q concentration note
current state · 2026-09-30written 2026-11-03
C48facthigh confidence
Q3 2026 revenue was $210.8M, up 55%: hardware fell 29% to $68.9M and cloud and other services rose 265% to $141.9M. GAAP gross margin was 20% and the operating loss $224.1M. On its core measures revenue was $230.9M, up 70%, the core gross margin 41% and the core operating margin -14%.
against this

Core revenue adds back $36.4M of warrant amortisation and removes $16.4M of pass-through revenue; nine-month stock pay was $541.7M.

historical · Q3 2026written 2026-11-12
C22factmedium confidence
SemiAnalysis said on 30 September 2026 that OpenAI's GPT-6.1 Sol Ultrafast runs on NVIDIA GPUs, not Cerebras; neither OpenAI nor Cerebras confirmed it.
against this
  • Cerebras capacity is still ramping and OpenAI may have used GPUs to launch quickly (S08).
  • Cerebras served GPT-5.6 Sol at 750 tokens a second in August (S03).
current state · 2026-09-30written 2026-11-03
C47inferencemedium confidence
On a three-stage fair value at an 8% required return Cerebras is worth about $32.83 a share against $101.08 at the 11 November close. Base case: trailing sales of $756M grow 60% and 60% over the next three years (the analysts' path, C46, held to at most 60% a year), then fade toward 4%, each year closing 30% of the gap, reaching it by year ten; the operating margin moves from -99.3% to 10.0% over five years, the analysts' year-three margin less stock pay; each added dollar of sales needs 50 cents of capital; a 21.0% tax rate; then 4% growth for ever at a 12% return on new capital; less $1.45B of debt and leases net of $7.80B of cash. Bear (half that growth, an 8.0% margin, new capital earning 8%) gives about $23.14; bull (the backlog on its filed schedule, revenue tripling then growing 150%, an 11.0% margin) about $143.65.
against this

The base case caps growth well below the backlog's own schedule: on that schedule (the bull case) the value is about $143.65, above the price. At a 12% required return the base is about $20.87.

  1. 2026-11-10editedre-valued on the owner's rules of 10 November: the return on new capital glides to its terminal level over years six to ten
  2. 2026-11-12editedre-struck after the Q3 print: trailing sales, the loss, cash, debt and shares from the Q3 10-Q at the 11 November close; the bull case follows the backlog's schedule
current state · 2026-11-12written 2026-11-09
C34inferencemedium confidence
If the backlog lands as scheduled, about $8.6B becomes revenue in the two years to September 2028 and about $11.5B in the two years after, near $5.7B a year by 2030, about 7.6 times trailing revenue.
against this
  • Timing can move at OpenAI's request (C06), and Cerebras has 700 MW of the capacity contracted so far (C27).
  • Revenue is not profit: the core gross margin was 41% in Q3 before data-centre wear is tested (C48, C23).
  1. 2026-11-12editedre-read on the Q3 10-Q's backlog schedule
current state · 2026-09-30written 2026-11-03
01

The company

When someone asks an AI model a question, a computer has to read the model's whole memory for every word it writes back. Graphics chips keep that memory off the chip, so they wait. Cerebras makes one chip from a whole silicon wafer and keeps the memory on it, so an independent tester measures it answering many times faster.

It sells these machines to buyers who run their own data centres, and rents their speed as a cloud service. OpenAI is now the largest renter.

The evidence under this - 4 claims
C01facthigh confidence
Cerebras sells AI computers built around one chip the size of a whole silicon wafer, as systems for a buyer's own data centre and as rented cloud capacity, and reports one segment.
current state · 2026-06-30written 2026-11-03
C18facthigh confidence
Artificial Analysis, an independent benchmarker, measures Cerebras serving gpt-oss-120b at 1,746 output tokens a second; the prospectus shows GPU clouds at 62 to 164 tokens a second on Llama 3.3 70B against Cerebras's 2,457.
against this
  • The GPU comparison mixes Cerebras's own measurements with the benchmark (S01).
  • SemiAnalysis said NVIDIA GPUs at low batch serve OpenAI's newest fast tier (C22).
current state · 2026-11-03written 2026-11-03
C19management claimhigh confidence
Cerebras keeps a model's weights in memory on the wafer itself, giving it 2,625 times the memory bandwidth of NVIDIA's B200, and avoids the HBM memory and CoWoS packaging in short supply.
against this
  • Large models must be split across several wafers, which adds complexity (S08).
  • SemiAnalysis long judged the design's weaknesses too hard to cover up until fast tokens found buyers (S07).
current state · 2026-05-14written 2026-11-03
C07facthigh confidence
In December 2025 OpenAI committed to buy 750 MW of Cerebras inference capacity on take-or-pay terms, deployed in tranches from 2026 to 2028 on three- or four-year terms extendable to five, with an option for 1.25 GW more by 2030.
against this
  • Revenue from OpenAI was $56.8M in Q2 2026, small beside the commitment, so delivery is still ahead (S02).
  • SemiAnalysis said OpenAI's newest fast tier runs on NVIDIA GPUs (C22).
  • On the Q3 call Cerebras said it is behind schedule on the first tranches, and either side may cancel a tranche delayed 90 days or more (C49).
  1. 2026-11-13editedre-read against the Q3 call: the finance chief restated the 750 MW take-or-pay terms and the 1.25 GW option; the delivery delay is recorded in C49
current state · 2026-11-12written 2026-11-03

How it makes money

Two ways. A buyer pays once for machines, which was most of the money until 2026. Or a customer rents capacity by the year: OpenAI pays for tranches of three or four years, so revenue arrives every quarter as data centres come online.

Renting is now most of sales and almost all of the backlog of orders. Cerebras also gave customers cheap shares as part of the deals, and that discount is taken off reported revenue until 2031.

The evidence under this - 5 claims
C03facthigh confidence
Q2 2026 revenue was $180.1M, up 74%: hardware fell 23% to $54.1M and cloud and other services rose 281% to $126.0M. GAAP gross margin was 14% and the operating loss $477.2M, including $377.0M of stock pay.
historical · Q2 2026written 2026-11-03
C07facthigh confidence
In December 2025 OpenAI committed to buy 750 MW of Cerebras inference capacity on take-or-pay terms, deployed in tranches from 2026 to 2028 on three- or four-year terms extendable to five, with an option for 1.25 GW more by 2030.
against this
  • Revenue from OpenAI was $56.8M in Q2 2026, small beside the commitment, so delivery is still ahead (S02).
  • SemiAnalysis said OpenAI's newest fast tier runs on NVIDIA GPUs (C22).
  • On the Q3 call Cerebras said it is behind schedule on the first tranches, and either side may cancel a tranche delayed 90 days or more (C49).
  1. 2026-11-13editedre-read against the Q3 call: the finance chief restated the 750 MW take-or-pay terms and the 1.25 GW option; the delivery delay is recorded in C49
current state · 2026-11-12written 2026-11-03
C06facthigh confidence
Remaining performance obligations were $26.1B at 30 September 2026, mostly OpenAI's; Cerebras expects about 33% to become revenue in the 24 months to September 2028 and 44% in months 25 to 48, against 22% and 43% at 30 June.
against this
  • Delivery timing may change at the customer's request, the filing says.
  • The figure excludes most future pass-through costs, which are billed on top.
  1. 2026-11-12editedre-read in the Q3 10-Q: the backlog grew and more of it is scheduled inside two years
current state · 2026-09-30written 2026-11-03
C10facthigh confidence
MBZUAI and G42, related parties in Abu Dhabi, were 62% and 24% of FY2025 revenue; in Q3 2026 Customer A, related to G42, was 38% and an unnamed Customer B 34%, and two customers held 80% of receivables at 30 September.
  1. 2026-11-12editedre-read in the Q3 10-Q concentration note
current state · 2026-09-30written 2026-11-03
C43facthigh confidence
Customer warrant assets of $1,092.0M at 30 September 2026 will reduce reported revenue through October 2031; $36.4M did so in Q3.
  1. 2026-11-12editedre-read in the Q3 10-Q balance sheet
current state · 2026-09-30written 2026-11-03
Where the money comes from, stream by stream
StreamShare of sales (Q2 2026)How it earns it
Cloud and other services70.0%Dedicated take-or-pay capacity, pay-as-you-go tokens, support and custom model work, plus data-centre costs passed through to OpenAI.
Hardware30.0%Cerebras systems sold for buyers' own data centres, mostly to G42 and MBZUAI, recognised on delivery.

Who runs it

Andrew Feldman, a co-founder, has run Cerebras since 2016, after selling his last chip company to AMD. He and Sean Lie, the other co-founder and technology chief, hold super-voting shares, and that class controls almost all the votes.

Their main award pays only if the company is worth $75B, $150B or $250B. Since the listing Lie and Feldman sold about $233M of shares outside pre-set trading plans, and no insider has bought on the market.

The evidence under this - 3 claims
C38facthigh confidence
Andrew Feldman, co-founder and chief executive, and Sean Lie, co-founder and chief technology officer, held 10.1M and 5.3M Class B shares before the IPO, 5.4% and 2.9% of the votes.
current state · 2026-03-31written 2026-11-03
C14facthigh confidence
The chief executive and chief technology officer were granted 9.0M performance units vesting at market values of $75B, $150B and $250B, beside 1.24M time-vesting units.
current state · 2026-05-14written 2026-11-03
C35facthigh confidence
Officers and directors made no open-market purchases since the IPO; the chief technology officer and chief executive sold about $233M of shares without a 10b5-1 flag from June to September, and later officer sales were under plans.
against this

The lock-up let insiders sell to cover tax on shares vesting at the IPO (S02); a missing flag is not proof of a view.

historical · 2026-05-14 to 2026-10-28written 2026-11-03
02

Why it wins, and who could beat it

Why customers come back

Speed is the reason customers pay, and speed now has a price. AI labs sell a fast tier of the same model for more: OpenAI charges about 6 times its standard price for its quickest answers.

Its design avoids the scarce memory graphics chips need, and OpenAI signing for 750 megawatts is outside proof a buyer values the speed. What the moat lacks is time: speed is a lead, not a lock-in, and a customer can move its fast tier when a rival matches it.

The evidence under this - 5 claims
C19management claimhigh confidence
Cerebras keeps a model's weights in memory on the wafer itself, giving it 2,625 times the memory bandwidth of NVIDIA's B200, and avoids the HBM memory and CoWoS packaging in short supply.
against this
  • Large models must be split across several wafers, which adds complexity (S08).
  • SemiAnalysis long judged the design's weaknesses too hard to cover up until fast tokens found buyers (S07).
current state · 2026-05-14written 2026-11-03
C18facthigh confidence
Artificial Analysis, an independent benchmarker, measures Cerebras serving gpt-oss-120b at 1,746 output tokens a second; the prospectus shows GPU clouds at 62 to 164 tokens a second on Llama 3.3 70B against Cerebras's 2,457.
against this
  • The GPU comparison mixes Cerebras's own measurements with the benchmark (S01).
  • SemiAnalysis said NVIDIA GPUs at low batch serve OpenAI's newest fast tier (C22).
current state · 2026-11-03written 2026-11-03
C20inferencemedium confidence
SemiAnalysis judges that labs now sell fast, priority, standard and batch tiers of one model, and that buyers paying more for speed is why OpenAI signed with Cerebras.
against this

OpenAI's Ultrafast tier reportedly runs at about 300 tokens a second on GPUs (S08), so a speed premium does not require Cerebras.

current state · 2026-05-13written 2026-11-03
C44factmedium confidence
OpenAI's Ultrafast tier is priced at $300 per million output tokens, about 6 times its standard price.
against this

The price is for OpenAI's own customers; what OpenAI pays Cerebras per megawatt is not disclosed (S02).

current state · 2026-09-30written 2026-11-03
C07facthigh confidence
In December 2025 OpenAI committed to buy 750 MW of Cerebras inference capacity on take-or-pay terms, deployed in tranches from 2026 to 2028 on three- or four-year terms extendable to five, with an option for 1.25 GW more by 2030.
against this
  • Revenue from OpenAI was $56.8M in Q2 2026, small beside the commitment, so delivery is still ahead (S02).
  • SemiAnalysis said OpenAI's newest fast tier runs on NVIDIA GPUs (C22).
  • On the Q3 call Cerebras said it is behind schedule on the first tranches, and either side may cancel a tranche delayed 90 days or more (C49).
  1. 2026-11-13editedre-read against the Q3 call: the finance chief restated the 750 MW take-or-pay terms and the 1.25 GW option; the delivery delay is recorded in C49
current state · 2026-11-12written 2026-11-03

The threat

NVIDIA, which sells most AI chips, paid about $2.9B for a licence to Groq's fast-answer design, the approach closest to Cerebras's. A research firm then reported that OpenAI's newest fast tier runs on NVIDIA chips, though neither company confirmed it.

The big cloud companies build their own AI chips too, and some of them, like Oracle, are also Cerebras's partners.

The evidence under this - 4 claims
C21facthigh confidence
NVIDIA paid $2,944M in the half to July 2026 for a non-exclusive licence of Groq's technology, an SRAM-based inference design.
historical · H1 FY2027written 2026-11-03
C22factmedium confidence
SemiAnalysis said on 30 September 2026 that OpenAI's GPT-6.1 Sol Ultrafast runs on NVIDIA GPUs, not Cerebras; neither OpenAI nor Cerebras confirmed it.
against this
  • Cerebras capacity is still ramping and OpenAI may have used GPUs to launch quickly (S08).
  • Cerebras served GPT-5.6 Sol at 750 tokens a second in August (S03).
current state · 2026-09-30written 2026-11-03
C24facthigh confidence
The prospectus names NVIDIA, AMD, Intel, chips built by hyperscalers, the AWS, Azure, Google and Oracle clouds, and CoreWeave as competitors, some of them also customers.
current state · 2026-05-14written 2026-11-03
C17facthigh confidence
Oracle says Cerebras inference on Oracle Cloud is planned for calendar 2027.
current state · 2026-10-27written 2026-11-03

The rivals: NVDA, CRWV, AMD, GOOG, AMZN, ORCL.

Who else does this, rival by rival
TickerRivalWhere it competesHow far alongPriced against it
NVDANVIDIA, including the Groq technology it licensedPriced against it: Not checkedCloud inference capacity, led by OpenAIOn saleNot checked
CRWVCoreWeave and other GPU cloudsPriced against it: Not checkedCloud inference capacity, led by OpenAIOn saleNot checked
AMD · GOOG · AMZNAMD and hyperscalers' own chipsPriced against it: Not checkedAI systems sold for buyers' own data centresOn saleNot checked
AMZN · ORCLAWS and Oracle serving fast tiers on their own or NVIDIA hardwarePriced against it: Not checkedHyperscaler and partner clouds: AWS, Oracle and paired inference with AMDOn saleNot checked
03

How big it could get

How big this could get

Cerebras, citing an industry research group, puts the market for answering AI questions at $66B in 2025, rising to $292B by 2029. A second forecaster sees all data-centre spending above $3 trillion a year by 2030.

Trailing sales are about 1% of that 2025 figure. Nobody measures the part that matters most here: how much buyers will pay extra for speed.

The evidence under this - 3 claims
C26management claimmedium confidence
Citing Bloomberg Intelligence, Cerebras puts the inference market it can address at $66B in 2025, rising to $292B by 2029.
against this

The figure is a vendor forecast chosen by the company, and it counts all inference, not the fast part Cerebras sells.

forwardwritten 2026-11-03
C25factmedium confidence
Dell'Oro forecasts data-centre capital spending above $3 trillion by 2030, with AI model builders and neoclouds growing near 60% a year.
against this

Dell'Oro says growth depends on investment holding up, power and supply (S10).

forwardwritten 2026-11-03
C42facthigh confidence
Trailing revenue to 30 September 2026 was $755.8M.
  1. 2026-11-12editedtrailing revenue rolled to Q3
current state · 2026-09-30written 2026-11-03

Who has signed

OpenAI has signed take-or-pay terms for 750 megawatts through 2028, and lent Cerebras about $1B to build it. The two Abu Dhabi buyers order machine by machine. Amazon signed a lease and a warrant that vests as it pays, Oracle agreed to buy machines and Microsoft cloud capacity once milestones are met; none of their sizes is disclosed.

The evidence under this - 6 claims
C07facthigh confidence
In December 2025 OpenAI committed to buy 750 MW of Cerebras inference capacity on take-or-pay terms, deployed in tranches from 2026 to 2028 on three- or four-year terms extendable to five, with an option for 1.25 GW more by 2030.
against this
  • Revenue from OpenAI was $56.8M in Q2 2026, small beside the commitment, so delivery is still ahead (S02).
  • SemiAnalysis said OpenAI's newest fast tier runs on NVIDIA GPUs (C22).
  • On the Q3 call Cerebras said it is behind schedule on the first tranches, and either side may cancel a tranche delayed 90 days or more (C49).
  1. 2026-11-13editedre-read against the Q3 call: the finance chief restated the 750 MW take-or-pay terms and the 1.25 GW option; the delivery delay is recorded in C49
current state · 2026-11-12written 2026-11-03
C06facthigh confidence
Remaining performance obligations were $26.1B at 30 September 2026, mostly OpenAI's; Cerebras expects about 33% to become revenue in the 24 months to September 2028 and 44% in months 25 to 48, against 22% and 43% at 30 June.
against this
  • Delivery timing may change at the customer's request, the filing says.
  • The figure excludes most future pass-through costs, which are billed on top.
  1. 2026-11-12editedre-read in the Q3 10-Q: the backlog grew and more of it is scheduled inside two years
current state · 2026-09-30written 2026-11-03
C09facthigh confidence
OpenAI lent Cerebras about $1.0B at 6%, repaid with service credits; $829.1M was outstanding at 30 September after $175.5M of credits, and it falls due at once if the contract ends other than through OpenAI's breach.
  1. 2026-11-12editedre-read in the Q3 10-Q
current state · 2026-09-30written 2026-11-03
C10facthigh confidence
MBZUAI and G42, related parties in Abu Dhabi, were 62% and 24% of FY2025 revenue; in Q3 2026 Customer A, related to G42, was 38% and an unnamed Customer B 34%, and two customers held 80% of receivables at 30 September.
  1. 2026-11-12editedre-read in the Q3 10-Q concentration note
current state · 2026-09-30written 2026-11-03
C15facthigh confidence
In June 2026 Cerebras gave Amazon a warrant for 2,696,678 shares at $100, vesting on Amazon's payments; the AWS lease had not started and earned nothing by 30 June.
current state · 2026-06-30written 2026-11-03
C16management claimmedium confidence
Cerebras signed agreements with Oracle to sell hardware and with Microsoft to provide cloud inference in the second half of 2026, the Microsoft one with milestones before service starts; management expects Cerebras on AWS Bedrock in Q1 2027, and the filing does not say how much of the $26.1B backlog any hyperscaler holds.
against this

Neither agreement's size is disclosed, the release says more will be shared next year, and the AWS lease had not started at 30 June (C15).

  1. 2026-11-12editedthe Q3 10-Q names Oracle and Microsoft agreements signed in the second half of 2026
forwardwritten 2026-11-03
04

The results

What growth costs

Before the listing, research spending grew slower than sales, so each new dollar of sales cost less to win. On the company's own measure the loss kept narrowing in 2026. The reported loss widened, but mostly because of share pay granted at the listing.

The evidence under this - 3 claims
C39facthigh confidence
In FY2025, R&D was 47.7% of revenue against 54.5% in FY2024, and the operating loss narrowed from 34.9% to 28.6% of revenue.
historical · FY2024 to FY2025written 2026-11-03
C04management claimhigh confidence
On its own core measures, which add back $44.3M of customer warrant amortisation and stock pay and remove pass-through data-centre costs, Q2 2026 gross margin was 41% and the operating loss $33.6M.
against this
  • GAAP gross margin was 14% in the same quarter (C03); the warrant amortisation is a real discount given to customers and recurs until 2031 (C08).
  • Core cloud gross margin includes depreciation struck on a three-to-five-year life that a short seller disputes (C23).
historical · Q2 2026written 2026-11-03
C03facthigh confidence
Q2 2026 revenue was $180.1M, up 74%: hardware fell 23% to $54.1M and cloud and other services rose 281% to $126.0M. GAAP gross margin was 14% and the operating loss $477.2M, including $377.0M of stock pay.
historical · Q2 2026written 2026-11-03

The last report (12 Nov)

Quarter ended June 2026, filed figures.

74%
Sales growth
against the same quarter a year earlier
-6.9%
Growth from the quarter before
against 12.8% the quarter before that
-265¢
Operating profit
per dollar of sales

In the third quarter sales rose 55% to $210.8M. On the company's own measure, which adds back share discounts given to customers, they were $230.9M, above the $214M to $216M guided, at a 41% margin. The year's sales guide rose to $940M to $960M, the backlog grew to $26.1B and Microsoft and Oracle signed deals of undisclosed size.

On the call the company said OpenAI's first batches of capacity are behind schedule because buildings are late. The backlog already subtracts the credits for that, and it expects to be back on schedule in the second half of 2027.

What it did to the thesisIt mostly supported the case: sales and margins beat the bar, the backlog grew and Microsoft signed. But the call said the OpenAI capacity the backlog rests on is running late, and how much the new deals are worth is still unknown.

The quarter in detail

Sales growth from the previous quarter, filed figures.

Earnings per share $-0.67 against $-0.15 expected.

The evidence under this - 7 claims
C48facthigh confidence
Q3 2026 revenue was $210.8M, up 55%: hardware fell 29% to $68.9M and cloud and other services rose 265% to $141.9M. GAAP gross margin was 20% and the operating loss $224.1M. On its core measures revenue was $230.9M, up 70%, the core gross margin 41% and the core operating margin -14%.
against this

Core revenue adds back $36.4M of warrant amortisation and removes $16.4M of pass-through revenue; nine-month stock pay was $541.7M.

historical · Q3 2026written 2026-11-12
C05management claimmedium confidence
After beating its Q3 guide, management guides Q4 2026 core revenue of $310M to $330M at a 47% to 49% core gross margin and a core operating margin of -1% to 1%, raises FY2026 core revenue to $940M to $960M from $880M to $890M, and expects to triple revenue in 2027.
against this
  • Core revenue adds back warrant discounts and removes pass-through revenue, so GAAP revenue runs lower: $210.8M against $230.9M in Q3 (C48).
  • SemiAnalysis reported OpenAI's newest fast tier running on NVIDIA GPUs, the customer the 2027 plan rests on (C22).
  1. 2026-11-12editedthe Q3 print beat the old guide; the guide was raised and a Q4 guide given
forwardwritten 2026-11-03
C06facthigh confidence
Remaining performance obligations were $26.1B at 30 September 2026, mostly OpenAI's; Cerebras expects about 33% to become revenue in the 24 months to September 2028 and 44% in months 25 to 48, against 22% and 43% at 30 June.
against this
  • Delivery timing may change at the customer's request, the filing says.
  • The figure excludes most future pass-through costs, which are billed on top.
  1. 2026-11-12editedre-read in the Q3 10-Q: the backlog grew and more of it is scheduled inside two years
current state · 2026-09-30written 2026-11-03
C16management claimmedium confidence
Cerebras signed agreements with Oracle to sell hardware and with Microsoft to provide cloud inference in the second half of 2026, the Microsoft one with milestones before service starts; management expects Cerebras on AWS Bedrock in Q1 2027, and the filing does not say how much of the $26.1B backlog any hyperscaler holds.
against this

Neither agreement's size is disclosed, the release says more will be shared next year, and the AWS lease had not started at 30 June (C15).

  1. 2026-11-12editedthe Q3 10-Q names Oracle and Microsoft agreements signed in the second half of 2026
forwardwritten 2026-11-03
C43facthigh confidence
Customer warrant assets of $1,092.0M at 30 September 2026 will reduce reported revenue through October 2031; $36.4M did so in Q3.
  1. 2026-11-12editedre-read in the Q3 10-Q balance sheet
current state · 2026-09-30written 2026-11-03
C49management claimmedium confidence
On the Q3 2026 call Cerebras said it is running behind schedule on OpenAI's initial capacity tranches because data-centre land, power and buildings are late; the $26.1B backlog is net of the delay credits it estimates OpenAI could claim; either side may cancel a tranche delayed 90 days or more, which it puts at under 5% of committed megawatts this year; no cancellation has been received, and it expects to be back on schedule in the second half of 2027.
against this

The size of the delay credits is not disclosed beyond 'relatively small', and the chief executive said OpenAI's own data centres are also behind schedule, so the recovery date rests on builders the company does not control.

forwardwritten 2026-11-13
C50management claimmedium confidence
On the Q3 2026 call the finance chief said core hardware gross margin fell to 31% on the last shipments under an old low-priced agreement while core cloud margin rose to 46.9%; that Q4 cloud revenue will be about flat, limited by data-centre capacity, with hardware near half of revenue in Q4 and Q1; that the Oracle hardware order, which he called large, is not in the $26.1B backlog; and the chief executive said paired inference with AMD GPUs is now in production.
against this

None of these figures is in the 10-Q or the release tables read in the first round, and the Oracle order's size is still undisclosed.

forwardwritten 2026-11-13
What management did with the money

The listing raised $6.2B, and first-half capital spending on data centres was $548.9M. Cerebras pays partly in its own shares: OpenAI's warrant could reach about 33M shares and Amazon's about 2.7M, both vesting as they buy.

Insiders in 2026Q3: 77 purchases and 266 sales, from their Form 4 filings.

The evidence under this - 4 claims
C11facthigh confidence
Cash, restricted cash and investments were $7,804.8M at 30 September 2026, down from $8,606.2M at 30 June as data-centre spending ran, with an undrawn $850M revolver.
  1. 2026-11-12editedre-read in the Q3 10-Q balance sheet
current state · 2026-09-30written 2026-11-03
C12facthigh confidence
In the first half of 2026 operations used $47.5M and capital spending was $548.9M; data-centre leases committed $690.3M, plus about $2.3B more signed with 2026 starts.
against this

The 10-Q gives no total cost for building 750 MW, so the remaining spend is unknown (S02).

historical · H1 2026written 2026-11-03
C08facthigh confidence
OpenAI holds a warrant for up to 33,445,026 Class N shares at almost no cost, vesting on the loan, capacity delivered and a market-value or payment threshold; 10,033,508 had vested and were exercised by July 2026.
current state · 2026-08-12written 2026-11-03
C15facthigh confidence
In June 2026 Cerebras gave Amazon a warrant for 2,696,678 shares at $100, vesting on Amazon's payments; the AWS lease had not started and earned nothing by 30 June.
current state · 2026-06-30written 2026-11-03
05

The price, and where we disagree

In short. Priced for perfection. The price is about 308% of what we think the business is worth, at an 8% yearly return.

Share price
$96.36
close,
Company value for each $1 of sales
23.3
its 4-year normal 849.9

The shares listed at $185 in May, opened at $350, and fell to $96.36 by early November after each of the first two results, the report on OpenAI's fast tier and a short seller's attack on the chips' life.

Third-quarter results on 12 November beat the company's guide and raised the year. The finance chief said the lock-up on insiders' shares had already ended on 9 November, freeing up to about 171M shares, and the fourth-quarter report, due by the end of March, is the next test.

The evidence under this - 7 claims
C28facthigh confidence
The IPO priced at $185 on 14 May 2026; the shares opened at $350, closed that day at $311.07, and closed at $96.36 on 2 November.
current state · 2026-11-02written 2026-11-03
C29facthigh confidence
The shares fell 19.6% the day after the first results, 11.8% after the second, 8.9% on the SemiAnalysis report and 5.7% on the Culper report.
historical · 2026-06-24 to 2026-10-28written 2026-11-03
C32facthigh confidence
Cerebras's finance chief said on the Q3 2026 call that the IPO lock-up releases were complete as of 9 November 2026; Cerebras had estimated up to about 171.1M shares released over the lock-up period.
against this

The thesis had read the lock-up as ending on the second trading day after the Q3 results; the call puts the end earlier, on 9 November, and the 10-Q was not reopened here to reconcile the two.

  1. 2026-11-13editedthe Q3 call says the lock-up releases were complete by 9 November, before the results, not two trading days after them
current state · 2026-11-12written 2026-11-03
C41facthigh confidence
Cerebras reported Q3 2026 results on 12 November 2026; Q4 2026 results are expected by the end of March 2027, a date not yet set.
  1. 2026-11-12editedthe Q3 results were reported on 12 November; the next dated report is Q4
current state · 2026-11-12written 2026-11-03
C48facthigh confidence
Q3 2026 revenue was $210.8M, up 55%: hardware fell 29% to $68.9M and cloud and other services rose 265% to $141.9M. GAAP gross margin was 20% and the operating loss $224.1M. On its core measures revenue was $230.9M, up 70%, the core gross margin 41% and the core operating margin -14%.
against this

Core revenue adds back $36.4M of warrant amortisation and removes $16.4M of pass-through revenue; nine-month stock pay was $541.7M.

historical · Q3 2026written 2026-11-12
C22factmedium confidence
SemiAnalysis said on 30 September 2026 that OpenAI's GPT-6.1 Sol Ultrafast runs on NVIDIA GPUs, not Cerebras; neither OpenAI nor Cerebras confirmed it.
against this
  • Cerebras capacity is still ramping and OpenAI may have used GPUs to launch quickly (S08).
  • Cerebras served GPT-5.6 Sol at 750 tokens a second in August (S03).
current state · 2026-09-30written 2026-11-03
C23factmedium confidence
Culper Research, short the stock, claimed on 28 October 2026 that Cerebras wafers last 1.4 to 2.3 years in service; Cerebras depreciates data-centre and computer equipment over 3 to 5 years.
against this
  • Culper's evidence is former employees and customers it does not name, reported second hand (S12).
  • Cerebras had not answered publicly when this was written.
current state · 2026-10-29written 2026-11-03
The betSafe, little upside
If it works
1.0×
from today's value, at the stated multiple
If it fails
keep 92%
of today's market value: the rest of the company, less its debt
How we value it
Method
Cash the business will produce, with any shortfall raised in new shares
Required return
8%
Long-run growth
4%
Moat
An average moat
BearBaseBull
Sales growth, years 1 to 330%, 30%60%, 60%200%, 150%
Normal operating margin8%10%11%
Return on new capital after year ten8%12%15%
Value a share$23.14$32.83$144
Price
$101
Priced on
11 Nov
Price as a share of the base value
308%
Yearly return the price offers on the base case
5.2%
Risk check: valued at 12%, the rate for a business not yet making a profit
$20.87

Where each figure comes from

FigureSourceWhere exactly
Sales, the last twelve monthsFiled reportRead by hand from the filed statements; the period it was read from was not recorded.
Operating margin todayFiled reportRead by hand from the filed statements; the period it was read from was not recorded.
Tax rateFiled reportRead by hand from the filed statements; the period it was read from was not recorded.
Sales per dollar of new capitalFiled reportRead by hand from the filed statements; the period it was read from was not recorded.
SharesFiled reportRead by hand from the filed statements; the period it was read from was not recorded.
Cash and investmentsFiled reportRead by hand from the filed statements; the period it was read from was not recorded.
DebtFiled reportRead by hand from the filed statements; the period it was read from was not recorded.
Sales growth, years 1 to 3Analysts' forecastsFMP's consensus for the years to 31 Dec 2026 and 31 Dec 2027, as it stood when the value was struck on 13 Nov 2026: the average of 7 to 8 analysts' revenue estimates, 60%, 60%, each year held to at most 60%. FMP publishes how many analysts, not who they are.
Normal operating marginAnalysts' forecastsFMP's consensus operating margin for the year to 31 Dec 2027, as it stood when the value was struck on 13 Nov 2026, the average of 8 analysts' estimates, less stock pay: 10% in the base case. FMP publishes how many analysts, not who they are.
Return on new capital after year tenOur method's ruleSet by our reading of the moat: the 8% rate with none, 12% with an average one, 15% with a strong one; the bear case earns the rate.
The bear and bull casesOur judgementOur reading of the evidence; each case below names it.
PriceMarket priceFMP's closing price on 11 Nov 2026.

What each case rests on

Bear caseGrowth runs at half the base path, 30% and 30%, the margin settles a fifth lower at 8.0%, and new capital earns only the 8% rate: the downturn test.

Base caseRevenue grows 60% and 60% over three years, the analysts' path (C46) held to at most 60% a year as on every thesis, then fades toward 4%, closing 30% of the gap each year; the margin moves from -99.3% to 10.0%, the analysts' year-three figure after stock pay. The backlog schedule runs well above this path, which is why the bull case follows it.

Bull caseThe backlog lands on the 10-Q's schedule: about $8.6B in the two years to September 2028 (C06, C34), roughly the analysts' uncapped path (C46), so revenue triples and then grows 150%; the margin reaches 11.0% and new capital earns 15%.

What moves the base value

LowerHigher
Growth in years 1 to 3, a quarter lower or higher$29.17$38.56
The normal margin, 15% lower or higher$29.07$36.59
Capital per dollar of growth, a quarter more or less$30.25$34.37
Return on new capital after year ten, at the rate or one step up$26.72$35.27
The rate, a point higher or lower$27.75$41.67
Long-run growth a point lower$30.19-

Value a share as growth and margin change

Normal margin ↓ · growth →50%75%100%125%150%
8%$24.94$25.89$27.81$31.11$38.57
9%$25.98$27.53$30.32$34.84$43.35
10%$27.02$29.17$32.83$38.56$48.23
11%$28.06$30.82$35.34$42.29$53.22
12%$29.10$32.46$37.84$46.02$58.30

Growth is shown as a share of the base case's. Green: the price is at or under 80% of that value. Red: at or over 120%.

How we read the multiple

There is no profit yet, and future sales sit mostly in one contract. On schedule, that backlog of orders alone brings sales near $5.7B a year by 2030.

The price is fair if OpenAI takes its capacity on time at a normal margin, and dear if either slips.

  1. Less its net cash, the company is worth 23.3 times last year's sales.
  2. Take a 9% yearly return, a common rule of thumb, and a sale at 4 times sales in ten years.
  3. Sales must then grow about 30.0% a year.

Against its own history. Its shares have traded only since May, so the older normal shown beside the price was struck before listing and says nothing about what buyers pay now.

The evidence under this - 5 claims
C33inferencemedium confidence
At $101.08 on 243.0M shares Cerebras is worth about $24,563M; less $6,976M of net cash, the company is worth about $17,587M, 23.3 times trailing revenue of $755.8M. At a 9% yearly return and a sale at 4 times revenue in ten years, revenue must grow about 30.0% a year; at 6 times, about 24.8%.
against this
  • The contracted backlog alone would reach several billion dollars a year by 2029 (C34), so early growth far exceeds 29.5%.
  • The test ignores the capital still to be spent on data centres (C12) and the OpenAI warrant's unvested shares (C45).
  1. 2026-11-12editedre-struck at the 11 November close on the Q3 balance sheet and trailing revenue
current state · 2026-11-12written 2026-11-03
C34inferencemedium confidence
If the backlog lands as scheduled, about $8.6B becomes revenue in the two years to September 2028 and about $11.5B in the two years after, near $5.7B a year by 2030, about 7.6 times trailing revenue.
against this
  • Timing can move at OpenAI's request (C06), and Cerebras has 700 MW of the capacity contracted so far (C27).
  • Revenue is not profit: the core gross margin was 41% in Q3 before data-centre wear is tested (C48, C23).
  1. 2026-11-12editedre-read on the Q3 10-Q's backlog schedule
current state · 2026-09-30written 2026-11-03
C42facthigh confidence
Trailing revenue to 30 September 2026 was $755.8M.
  1. 2026-11-12editedtrailing revenue rolled to Q3
current state · 2026-09-30written 2026-11-03
C06facthigh confidence
Remaining performance obligations were $26.1B at 30 September 2026, mostly OpenAI's; Cerebras expects about 33% to become revenue in the 24 months to September 2028 and 44% in months 25 to 48, against 22% and 43% at 30 June.
against this
  • Delivery timing may change at the customer's request, the filing says.
  • The figure excludes most future pass-through costs, which are billed on top.
  1. 2026-11-12editedre-read in the Q3 10-Q: the backlog grew and more of it is scheduled inside two years
current state · 2026-09-30written 2026-11-03
C47inferencemedium confidence
On a three-stage fair value at an 8% required return Cerebras is worth about $32.83 a share against $101.08 at the 11 November close. Base case: trailing sales of $756M grow 60% and 60% over the next three years (the analysts' path, C46, held to at most 60% a year), then fade toward 4%, each year closing 30% of the gap, reaching it by year ten; the operating margin moves from -99.3% to 10.0% over five years, the analysts' year-three margin less stock pay; each added dollar of sales needs 50 cents of capital; a 21.0% tax rate; then 4% growth for ever at a 12% return on new capital; less $1.45B of debt and leases net of $7.80B of cash. Bear (half that growth, an 8.0% margin, new capital earning 8%) gives about $23.14; bull (the backlog on its filed schedule, revenue tripling then growing 150%, an 11.0% margin) about $143.65.
against this

The base case caps growth well below the backlog's own schedule: on that schedule (the bull case) the value is about $143.65, above the price. At a 12% required return the base is about $20.87.

  1. 2026-11-10editedre-valued on the owner's rules of 10 November: the return on new capital glides to its terminal level over years six to ten
  2. 2026-11-12editedre-struck after the Q3 print: trailing sales, the loss, cash, debt and shares from the Q3 10-Q at the 11 November close; the bull case follows the backlog's schedule
current state · 2026-11-12written 2026-11-09
Priced against its rivals

Its rivals have not yet been checked against their own filings, so there is no group to price it against; the price is judged against its own history instead.

The bet: what it pays if it works

The bet is on Cerebras renting its speed through the big cloud companies, starting with Amazon and Oracle in 2027.

If both run at a solid pace, the business would be worth about what the whole company costs today, so winning adds little. If neither pays, holders keep most of the value, because the OpenAI cloud and the cash already account for it.

Hyperscaler and partner clouds: AWS, Oracle and paired inference with AMD revenue in two to three yearsDollars of yearly hyperscaler revenue neededAt the stated multiple
$100.00Mabout 100.0M at 100% kept0.9×
$500.00M (the big win)about 500.0M at 100% kept1.0×
$1.00Babout 1.0B at 100% kept1.2×

To pay 2× (a Fair bet) it would need about $3.41B of Hyperscaler and partner clouds: AWS, Oracle and paired inference with AMD revenue (about 3.4B dollars of yearly hyperscaler revenue); to pay 5× (a Cheap bet), about $12.57B of Hyperscaler and partner clouds: AWS, Oracle and paired inference with AMD revenue (about 12.6B dollars of yearly hyperscaler revenue), at the same multiple.

Each row is Hyperscaler and partner clouds: AWS, Oracle and paired inference with AMD alone, plus the rest of the company (Cloud inference capacity, led by OpenAI) counted at about $13.50B.

The steps it has to take

  • Done Amazon warrant and AWS lease signed
  • Done Paired inference with AMD in production
  • In progress Cerebras live on AWS Bedrock due 2027-03-31
  • Not started First revenue from other hyperscalers due 2027-06-30
  • Not started Cerebras inference on Oracle Cloud due 2027-12-31
The detail behind the label
Rating
Expensive
the big win's payoff against the 2× and 5× lines
Cheap bet at about
≈ $45.70
53% below the price (close, )
The option
1.4×
what Hyperscaler and partner clouds: AWS, Oracle and paired inference with AMD is worth if it works, against what the price pays for it
The evidence under this - 5 claims
C15facthigh confidence
In June 2026 Cerebras gave Amazon a warrant for 2,696,678 shares at $100, vesting on Amazon's payments; the AWS lease had not started and earned nothing by 30 June.
current state · 2026-06-30written 2026-11-03
C16management claimmedium confidence
Cerebras signed agreements with Oracle to sell hardware and with Microsoft to provide cloud inference in the second half of 2026, the Microsoft one with milestones before service starts; management expects Cerebras on AWS Bedrock in Q1 2027, and the filing does not say how much of the $26.1B backlog any hyperscaler holds.
against this

Neither agreement's size is disclosed, the release says more will be shared next year, and the AWS lease had not started at 30 June (C15).

  1. 2026-11-12editedthe Q3 10-Q names Oracle and Microsoft agreements signed in the second half of 2026
forwardwritten 2026-11-03
C17facthigh confidence
Oracle says Cerebras inference on Oracle Cloud is planned for calendar 2027.
current state · 2026-10-27written 2026-11-03
C33inferencemedium confidence
At $101.08 on 243.0M shares Cerebras is worth about $24,563M; less $6,976M of net cash, the company is worth about $17,587M, 23.3 times trailing revenue of $755.8M. At a 9% yearly return and a sale at 4 times revenue in ten years, revenue must grow about 30.0% a year; at 6 times, about 24.8%.
against this
  • The contracted backlog alone would reach several billion dollars a year by 2029 (C34), so early growth far exceeds 29.5%.
  • The test ignores the capital still to be spent on data centres (C12) and the OpenAI warrant's unvested shares (C45).
  1. 2026-11-12editedre-struck at the 11 November close on the Q3 balance sheet and trailing revenue
current state · 2026-11-12written 2026-11-03
C05management claimmedium confidence
After beating its Q3 guide, management guides Q4 2026 core revenue of $310M to $330M at a 47% to 49% core gross margin and a core operating margin of -1% to 1%, raises FY2026 core revenue to $940M to $960M from $880M to $890M, and expects to triple revenue in 2027.
against this
  • Core revenue adds back warrant discounts and removes pass-through revenue, so GAAP revenue runs lower: $210.8M against $230.9M in Q3 (C48).
  • SemiAnalysis reported OpenAI's newest fast tier running on NVIDIA GPUs, the customer the 2027 plan rests on (C22).
  1. 2026-11-12editedthe Q3 print beat the old guide; the guide was raised and a Q4 guide given
forwardwritten 2026-11-03

Where we disagree with the market

The market believes

Analysts, all positive, count the backlog of orders; short sellers, holding a third of the shares that trade, count one customer, wearing chips and NVIDIA.

We believe

Both are partly right: the contract is real, one buyer carries it, and today's price is about what it would earn.

The bulls may be missing that a fast tier can move: if NVIDIA's chips match the speed, OpenAI has less reason to take every tranche. The bears may be missing that the contract is take-or-pay and OpenAI lent Cerebras the money to build it.

The evidence against our middle view is the margin. If the cloud earns its 41% after the chips are paid for, the price is cheap; if the chips wear out early, it is dear.

What analysts say, and the evidence
Analysts positive
100%
of 11, 1 Nov
Four months earlier
91%
of 11, 1 Jul
The evidence under this - 7 claims
C30facthigh confidence
All 11 analysts in FMP's series rated the shares positive on 1 November 2026, against 90.9% of 11 in July.
current state · 2026-11-01written 2026-11-03
C31facthigh confidence
Short interest was 19.9M shares, 32.6% of the float, at the 15 October settlement, up 40.7% from the one before.
current state · 2026-10-15written 2026-11-03
C33inferencemedium confidence
At $101.08 on 243.0M shares Cerebras is worth about $24,563M; less $6,976M of net cash, the company is worth about $17,587M, 23.3 times trailing revenue of $755.8M. At a 9% yearly return and a sale at 4 times revenue in ten years, revenue must grow about 30.0% a year; at 6 times, about 24.8%.
against this
  • The contracted backlog alone would reach several billion dollars a year by 2029 (C34), so early growth far exceeds 29.5%.
  • The test ignores the capital still to be spent on data centres (C12) and the OpenAI warrant's unvested shares (C45).
  1. 2026-11-12editedre-struck at the 11 November close on the Q3 balance sheet and trailing revenue
current state · 2026-11-12written 2026-11-03
C34inferencemedium confidence
If the backlog lands as scheduled, about $8.6B becomes revenue in the two years to September 2028 and about $11.5B in the two years after, near $5.7B a year by 2030, about 7.6 times trailing revenue.
against this
  • Timing can move at OpenAI's request (C06), and Cerebras has 700 MW of the capacity contracted so far (C27).
  • Revenue is not profit: the core gross margin was 41% in Q3 before data-centre wear is tested (C48, C23).
  1. 2026-11-12editedre-read on the Q3 10-Q's backlog schedule
current state · 2026-09-30written 2026-11-03
C22factmedium confidence
SemiAnalysis said on 30 September 2026 that OpenAI's GPT-6.1 Sol Ultrafast runs on NVIDIA GPUs, not Cerebras; neither OpenAI nor Cerebras confirmed it.
against this
  • Cerebras capacity is still ramping and OpenAI may have used GPUs to launch quickly (S08).
  • Cerebras served GPT-5.6 Sol at 750 tokens a second in August (S03).
current state · 2026-09-30written 2026-11-03
C23factmedium confidence
Culper Research, short the stock, claimed on 28 October 2026 that Cerebras wafers last 1.4 to 2.3 years in service; Cerebras depreciates data-centre and computer equipment over 3 to 5 years.
against this
  • Culper's evidence is former employees and customers it does not name, reported second hand (S12).
  • Cerebras had not answered publicly when this was written.
current state · 2026-10-29written 2026-11-03
C06facthigh confidence
Remaining performance obligations were $26.1B at 30 September 2026, mostly OpenAI's; Cerebras expects about 33% to become revenue in the 24 months to September 2028 and 44% in months 25 to 48, against 22% and 43% at 30 June.
against this
  • Delivery timing may change at the customer's request, the filing says.
  • The figure excludes most future pass-through costs, which are billed on top.
  1. 2026-11-12editedre-read in the Q3 10-Q: the backlog grew and more of it is scheduled inside two years
current state · 2026-09-30written 2026-11-03
06

What could go wrong, and how we would know

The risks

  • One buyer carries it. Most of the backlog of orders is OpenAI's, its first batches of capacity are running late, and before that two related Abu Dhabi buyers paid 86% of sales.
  • NVIDIA closes the speed gap. NVIDIA licensed Groq's fast design and reportedly serves OpenAI's newest fast tier, which would leave Cerebras's speed less special.
  • The chips may wear out early. A short seller claims the wafers last under two and a half years in service; the company writes them off over three to five.
  • Margins nobody has seen. The reported gross margin was 14%; the 41% figure removes share discounts to customers that recur until 2031.
  • More shares are coming. Unvested warrants and founders' units could add about 15% to the share count, and the lock-up on insiders' shares ended on 9 November.
The evidence under this - 10 claims
C06facthigh confidence
Remaining performance obligations were $26.1B at 30 September 2026, mostly OpenAI's; Cerebras expects about 33% to become revenue in the 24 months to September 2028 and 44% in months 25 to 48, against 22% and 43% at 30 June.
against this
  • Delivery timing may change at the customer's request, the filing says.
  • The figure excludes most future pass-through costs, which are billed on top.
  1. 2026-11-12editedre-read in the Q3 10-Q: the backlog grew and more of it is scheduled inside two years
current state · 2026-09-30written 2026-11-03
C10facthigh confidence
MBZUAI and G42, related parties in Abu Dhabi, were 62% and 24% of FY2025 revenue; in Q3 2026 Customer A, related to G42, was 38% and an unnamed Customer B 34%, and two customers held 80% of receivables at 30 September.
  1. 2026-11-12editedre-read in the Q3 10-Q concentration note
current state · 2026-09-30written 2026-11-03
C22factmedium confidence
SemiAnalysis said on 30 September 2026 that OpenAI's GPT-6.1 Sol Ultrafast runs on NVIDIA GPUs, not Cerebras; neither OpenAI nor Cerebras confirmed it.
against this
  • Cerebras capacity is still ramping and OpenAI may have used GPUs to launch quickly (S08).
  • Cerebras served GPT-5.6 Sol at 750 tokens a second in August (S03).
current state · 2026-09-30written 2026-11-03
C21facthigh confidence
NVIDIA paid $2,944M in the half to July 2026 for a non-exclusive licence of Groq's technology, an SRAM-based inference design.
historical · H1 FY2027written 2026-11-03
C23factmedium confidence
Culper Research, short the stock, claimed on 28 October 2026 that Cerebras wafers last 1.4 to 2.3 years in service; Cerebras depreciates data-centre and computer equipment over 3 to 5 years.
against this
  • Culper's evidence is former employees and customers it does not name, reported second hand (S12).
  • Cerebras had not answered publicly when this was written.
current state · 2026-10-29written 2026-11-03
C04management claimhigh confidence
On its own core measures, which add back $44.3M of customer warrant amortisation and stock pay and remove pass-through data-centre costs, Q2 2026 gross margin was 41% and the operating loss $33.6M.
against this
  • GAAP gross margin was 14% in the same quarter (C03); the warrant amortisation is a real discount given to customers and recurs until 2031 (C08).
  • Core cloud gross margin includes depreciation struck on a three-to-five-year life that a short seller disputes (C23).
historical · Q2 2026written 2026-11-03
C43facthigh confidence
Customer warrant assets of $1,092.0M at 30 September 2026 will reduce reported revenue through October 2031; $36.4M did so in Q3.
  1. 2026-11-12editedre-read in the Q3 10-Q balance sheet
current state · 2026-09-30written 2026-11-03
C45inferencemedium confidence
About 26.1M contingently issuable warrant shares, mostly OpenAI's and Amazon's, plus 9.0M executive performance units, together about 14.4% of the 243.0M shares outstanding.
against this

Some OpenAI tranches vest only if OpenAI buys the optional 1.25 GW (S02), and the executive units need a market value far above today's.

  1. 2026-11-12editedre-read on the Q3 10-Q's contingently issuable shares
current state · 2026-11-05written 2026-11-03
C32facthigh confidence
Cerebras's finance chief said on the Q3 2026 call that the IPO lock-up releases were complete as of 9 November 2026; Cerebras had estimated up to about 171.1M shares released over the lock-up period.
against this

The thesis had read the lock-up as ending on the second trading day after the Q3 results; the call puts the end earlier, on 9 November, and the 10-Q was not reopened here to reconcile the two.

  1. 2026-11-13editedthe Q3 call says the lock-up releases were complete by 9 November, before the results, not two trading days after them
current state · 2026-11-12written 2026-11-03
C49management claimmedium confidence
On the Q3 2026 call Cerebras said it is running behind schedule on OpenAI's initial capacity tranches because data-centre land, power and buildings are late; the $26.1B backlog is net of the delay credits it estimates OpenAI could claim; either side may cancel a tranche delayed 90 days or more, which it puts at under 5% of committed megawatts this year; no cancellation has been received, and it expects to be back on schedule in the second half of 2027.
against this

The size of the delay credits is not disclosed beyond 'relatively small', and the chief executive said OpenAI's own data centres are also behind schedule, so the recovery date rests on builders the company does not control.

forwardwritten 2026-11-13
What we still do not know
  • Whether OpenAI's newest fast tier really runs on NVIDIA chips, and whether that changes how much Cerebras capacity it takes.
  • How long the wafer systems last in service, which sets what the cloud really earns.
  • Whether any big cloud company signs paid capacity, rather than a lease and a warrant.
  • What the six new deals above $30M each are worth, since no customer has disclosed one.
The evidence under this - 4 claims
C22factmedium confidence
SemiAnalysis said on 30 September 2026 that OpenAI's GPT-6.1 Sol Ultrafast runs on NVIDIA GPUs, not Cerebras; neither OpenAI nor Cerebras confirmed it.
against this
  • Cerebras capacity is still ramping and OpenAI may have used GPUs to launch quickly (S08).
  • Cerebras served GPT-5.6 Sol at 750 tokens a second in August (S03).
current state · 2026-09-30written 2026-11-03
C23factmedium confidence
Culper Research, short the stock, claimed on 28 October 2026 that Cerebras wafers last 1.4 to 2.3 years in service; Cerebras depreciates data-centre and computer equipment over 3 to 5 years.
against this
  • Culper's evidence is former employees and customers it does not name, reported second hand (S12).
  • Cerebras had not answered publicly when this was written.
current state · 2026-10-29written 2026-11-03
C16management claimmedium confidence
Cerebras signed agreements with Oracle to sell hardware and with Microsoft to provide cloud inference in the second half of 2026, the Microsoft one with milestones before service starts; management expects Cerebras on AWS Bedrock in Q1 2027, and the filing does not say how much of the $26.1B backlog any hyperscaler holds.
against this

Neither agreement's size is disclosed, the release says more will be shared next year, and the AWS lease had not started at 30 June (C15).

  1. 2026-11-12editedthe Q3 10-Q names Oracle and Microsoft agreements signed in the second half of 2026
forwardwritten 2026-11-03
C37management claimmedium confidence
In Q2 2026 Cerebras signed six deals above $30M each, with customers including Cognition, Lovable, Figma, Block, AlphaSense, GSK and CrowdStrike.
against this

None of these customers filed or announced a contract value, and none appears among the 10% customers (C10).

historical · Q2 2026written 2026-11-03

What would change our view

  1. Unattractive→NeutralThe price falls back within 20% of fair value

    It needs:

    • ○The share price falls to about $39.40, back within 20% of what we think the business is worth.about ≈$18.04 today, 81% below the last close
  2. Unattractive→AttractiveThe price falls 20% under fair value

    It needs:

    • ○The share price falls to about $26.26, 80% of what we think the business is worth.about ≈$0.13 today, 100% below the last close
  3. Unattractive→AttractiveProof the backlog is not one customer

    It needs all of:

    • ○A customer other than OpenAI signs $2B or more of take-or-pay capacity.watched as it happens
    • ○The company's own gross margin stays at 38% or more for two quarters.checked at the next results, due by 31 Mar
We would also be wrong if
  • OpenAI cuts, delays or renegotiates its 750 megawatts.
  • Cerebras shortens the life it writes its data-centre systems off over to under three years.
  • No big cloud company or second large customer joins the backlog of orders by the end of 2027.

What is coming

  • 16 November 2026 Lock-up ends: How many of the roughly 171M locked shares holders sell.
Every dated event ahead
  • 16 November 2026 Lock-up ends: How many of the roughly 171M locked shares holders sell.
  • 31 March 2027 AWS Bedrock goes live with Cerebras: The first hyperscaler revenue, and whether Amazon's warrant starts vesting.
  • 31 March 2027 Q4 2026 results: FY2026 against the raised $940M to $960M core guide, the 47% to 49% core margin, and the 2027 plan restated.
Since we wrote this
Written
3 Nov
Price the thesis read
$96.36
2 Nov
  • 13 NovEarningsCall says OpenAI capacity is late; Unattractive heldThe finance chief said the first OpenAI tranches are behind schedule on late buildings, with delay credits already deducted from the backlog. Our value does not rest on that schedule, so the view holds.
  • 12 NovEarningsQ3 beat and raised; Unattractive held, conviction cut to mediumRevenue and margins beat the bar and Microsoft signed, but our capped value sits far below the price; only the backlog landing on schedule would reach it.
  • 11 NovThesis reviewUnattractive, high conviction: the view holds even in our best caseConviction re-graded on the new anchors: how robust the view is, not whether anything is still open.
  • 10 NovNewsBilt picks Cerebras for its AI assistant, with no size or price disclosed. View heldBilt selects Cerebras to power its Neighborhood OS and Concierge assistant.
  • 9 NovThesis reviewUnattractive: priced at 221% of a value near $40.51Re-valued in three stages at an 8% return, the method now drawn on every thesis: the price sits beyond the 20% margin of safety.
  • 3 NovThesis writtenNeutral, low convictionWritten with the shares at $96.36, below the $185 IPO price, nine days before the Q3 2026 results and the end of the lock-up.

Sources 19

Where this came from

Sources - 19

Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.

  1. S01primarycompany filing2026-05-14
    business, competition, market sizing, FY2024 and FY2025 statements, customer concentration, useful lives, share classes, ownership, executive grants
    backsC01C02C07C10C13C14C18C19C23C24C26C33C38C39C42
  2. S02primarycompany filing2026-08-12
    results, OpenAI MRA, working capital loan, warrants, remaining performance obligations, customer concentration, cash, leases, lock-up
    backsC01C03C04C07C08C12C15C28C32C40
  3. S03primarycompany release2026-08-12
    core measures, Q3 and FY2026 guidance, capacity and customer highlights
    We have made rapid progress in key areas required to deliver exceptional growth against our remaining performance obligations of $25.4 billion, and plan to more than triple revenue in 2027.
    backsC04C19C27C37
  4. S04primarytranscript2026-08-12
    the 2027 plan, data-centre capacity, hyperscaler timing, disaggregated inference, analyst questions on concentration
    And with all of this progress, I think it's important to keep in mind that our $25 billion in RPO does not reflect any backlog of business from AWS or any other hyperscaler at this time.
    backsC16C36C37
  5. S05secondaryvendor data2026-11-03
    Research pack, FMP and FINRA market data
    content/CBRS/research-pack.json
    price history, analyst split, short interest, earnings date, insider feed
    Cerebras Systems Inc. is a leading innovator in artificial intelligence infrastructure.
    backsC28C29C30C31C32
  6. S06primaryindustry data2026-11-03
    measured output speed and price on Cerebras
    For output speed, the fastest models are gpt-oss-120b (low) (1,746 t/s) and gpt-oss-120b (high) (1,684 t/s).
    backsC18
  7. S07secondaryanalyst2026-05-13
    why buyers pay for speed; Nvidia's Groq deal described as SRAM-based
    with frontier labs releasing fast, priority, standard and batch tiers of the same model weights, the world has revealed their preference for fast tokens with their wallets.
    backsC20C21
  8. S08secondarynews2026-09-30
    the SemiAnalysis claim, Ultrafast pricing, OpenAI's mixed fleet
    OpenAI has not said publicly which hardware serves the 6.1 Sol Ultrafast tier, and Cerebras has not commented on the claim at the time of writing.
    backsC22C44
  9. S09primarycompany filing2026-08-26
    the payment for the non-exclusive Groq licence
    backsC21
  10. S10primaryindustry data2026-08-18
    data center capex outlook and the growth of AI-specialised clouds
    The newly added AI-specialized cloud segment, comprising AI model builders and neocloud service providers, is projected to grow at nearly a 60 percent CAGR, outpacing the growth of other customer segments.
    backsC25
  11. S11primarycompany release2026-10-27
    the Oracle offering and its date
    Cerebras Inference on OCI is planned to be available in calendar year 2027, giving OCI customers another option for AI applications where speed is critical.
    backsC17
  12. S12secondarynews2026-10-29
    Culper Research's short claims, as reported
    In draft IPO filings, Cerebras had said that its chips could last between three and five years, but Culper said the wafer-scale chips last just 1.4 to 2.3 years.
    backsC23
  13. S13primarycompany filing2026-10-28
    open-market sales and the Rule 10b5-1 flag
    backsC35
  14. S14primarycompany filing2026-11-09
    trailing revenue, operating profit, stock pay, tax, debt, cash and the diluted share count
    read, but no claim rests on it
  15. S15secondaryvendor data2026-11-09
    the analysts' average revenue and operating profit for the next three fiscal years, and the close
    backsC46
  16. S16primarycompany filing2026-11-12
    results, remaining performance obligations and their schedule, customer concentration, the OpenAI loan, warrants, cash, leases, the Oracle and Microsoft agreements, shares outstanding at 5 November
    backsC06C09C10C11C13C16C33C34C42C43C45C47C48
  17. S17primarycompany filing2026-11-12
    core revenue and margins, the Q4 and FY2026 core guide, 700 MW of capacity, the hyperscaler cloud deal
    backsC05C11C16C27C41C48
  18. S18secondaryvendor data2026-11-12
    the close and the analysts' average revenue and operating profit for fiscal 2026 to 2028
    backsC33C47
  19. S19primarytranscript2026-11-12
    the OpenAI delivery delay, delay credits and cancellation terms, the lock-up end, segment core margins, the Q4 mix, Oracle outside the backlog, disaggregated inference with AMD in production, the CS-5 schedule; questions from Barclays, Morgan Stanley, Citi, Needham and TD Cowen
    After signing the agreement with OpenAI, it took us longer than we would have liked to find contracts for and bring online the substantial increase in data center capacity required to meet the aggressive requirements in the agreement. And as such, we find ourselves running behind schedule on initial tranches.
    backsC07C32C36C49C50
Rendered from content/CBRS/thesis.json · thesis reviewed 2026-11-13. This page explains a company; it does not say what to pay, when to buy or how much.