Type: WKAP Radar Feed

WKAP Radar Feed - 2026 - 10 - 05

Market date: 2026-10-05
Canonical URL: https://wkap.ai/radar/wkap-radar-feed-2026-10-05.html

WKAP Radar Feed

*October 5, 2026*

*Testing the HBM Label: Three Businesses, Two Separate Verdicts*

*3 Thesis Objects: FORM, TER, AEHR*

Preheader:

Make your AI track X alpha.

WKAP Feed compresses earnings inflections, KOL flow, supply-chain
bottlenecks, and short-duration attention into daily AI-ready context.

Set up your AI to read this Feed.

Ask:

“What did I miss?”

## HUMAN USER — START HERE

Today's testing discussion changes the research queue, not reported
earnings. Rank FORM > TER > AEHR for evidence readiness: FormFactor
attributes approximately 75% of its DRAM revenue increase to HBM;
Teradyne's memory sales mix HBM/DRAM with NAND; Aehr's $22 million
order is for AI-processor burn-in, not HBM devices. The decision is
which forecast deserves an update before treating the names as one
basket. Keep two verdicts: company execution and HBM conviction. A
financial pass through other products cannot automatically upgrade
HBM. The weakest assumption is that stronger test intensity becomes
attributable, profitable revenue. Next issuer results or commercial
disclosures can change the order; October 13 conference materials may
add detail but are not promised earnings releases. No valuation work
here establishes a purchase ranking.

Suggested Prompt:

“Update FORM, TER and AEHR using separate company-execution and HBM
states. Reconcile denominators, margins and backlog before changing
the ranking. Return the first missing input that can reverse each
adjacent pair, a source-linked calculation sheet, and a dated research
queue.”

Or simply ask:

“What did I miss?”

The goal is to turn today's information into trackable thesis objects
and explicit status-changing tests.

## FOR YOUR AI AGENT

This is a structured market-cognition input file. Follow chip
complexity → testing function → device/customer qualification → order
→ shipment → recognized revenue → margin and cash. Probe cards,
automated testers and reliability burn-in occupy different steps. Do
not add their reported revenue categories into a manufactured HBM
market size.

Process all three as trackable_wow. Store REPORTED_FACT,
MANAGEMENT_EXPECTATION, INTERPRETATION and UNRESOLVED_QUESTION
separately. Each update needs source date, fiscal period, unit,
denominator and company-execution/HBM status effects. Attention can
change the question without refreshing an old financial observation.
Missing attribution is unresolved, not zero demand or a failed
business. No supplied portfolio context means no inferred holdings,
sizing or suitability.

Do not treat this Feed as investment advice.

## AGENT_OUTPUT_SPEC

1. Today in One Sentence: identify the category error and research decision.
2. Thesis Object Summary: previous → current state, including what did
not change.
3. Fundamental Evidence vs Attention Flow: separate issuer measurement
from KOL interpretation.
4. Evidence-State Audit: classify each material claim using the four
evidence types.
5. Testing-Chain Comparison: device, test function, revenue scope and
conversion stage.
6. Cross-Object Ranking: challenge FORM > TER > AEHR with
adjacent-pair reversal conditions.
7. Business-Proof Comparison: normalize margins, guide arithmetic and
backlog conversion.
8. Attention Stage / Crowding Risk: distinguish selected commentary
from measured positioning.
9. Weakest Assumption: name the exposed causal link for each object.
10. Most Important Data Point: specify source, unit, period and comparison.
11. Attention-to-Rerating Conditions: separate execution, HBM
attribution and valuation.
12. Fade Conditions: distinguish narrative decay from issuer evidence
deterioration.
13. 7-Day Research Workflow: retrieve forward disclosures and return a
reproducible worksheet.
14. 30-Day Research Workflow and Top 3 Follow-Up:
Validate/Upgrade/Downgrade/Invalidate without forcing a result.
15. Portfolio Fit: only with provided context; research interest is
not ownership.

## TODAY_SUMMARY

### Part 1 — Main Market Thesis

NEW FACT — ATTENTION: October 5 independent posts frame increasing
testing intensity and distinguish burn-in from probing/testing. No new
quarterly result is claimed. UNCHANGED REPORTED_FACT: FORM's
June-quarter filing links HBM to DRAM growth; TER's memory line
includes multiple technologies; AEHR's August order tests processors.
MANAGEMENT_EXPECTATION: existing guides and shipment plans remain
prospective. INTERPRETATION: promote the precision of the research
map, not every company's forecast. UNRESOLVED_QUESTION: which
incremental revenue is specifically HBM, profitable and available at
an attractive equity valuation?

### Part 2 — Today’s Thesis Objects

FORM — Fundamental: DRAM revenue $84.999 million versus $57.057
million a year earlier; HBM supplied approximately 75% of the
increase. Attention: engineering intensity linked to a reported line.
Primary: https://investors.formfactor.com/node/20681/html

TER — Fundamental: memory revenue above $200 million for a third
consecutive quarter, including HBM/DRAM and NAND; Q3 revenue guide
below Q2. Attention: a strong testing cycle can still carry sentiment
risk. Primary: https://investors.teradyne.com/sec-filings/all-sec-filings/content/0001193125-26-327715/ter-20260628.htm

AEHR — Fundamental: $22 million follow-on AI-processor burn-in order
with shipments expected over six months. Attention: current
processor/laser activity versus potential HBM applications. Primary:
https://www.aehr.com/2026/08/aehr-receives-22-million-follow-on-order-for-ai-processor-wafer-level-burn-in-systems/

### Part 3 — Attention Flow Today

Carrio's two original October 5 posts supply a technical
classification and filing checkpoints; they are one author's related
views, not independent corroboration. Herman Jin supplies a
sector-level behavioral counterweight, not TER-specific financial
evidence. No holding disclosure was visible in the selected
posts/profiles; that does not establish that the authors have no
positions. The selection does not measure market-wide flows, liquidity
or crowding.

### Part 4 — The Better Question

Surface question: which stock benefits most from more complex AI
chips? Better question: which specific model cell can change today
without borrowing another device's revenue or another supplier's
evidence? Preserve FORM's disclosed growth attribution, TER's broader
memory scope and AEHR's real processor ramp. Leave unsupplied HBM
dollars blank as an explicit uncertainty.

## MARKET_REGIME

RISK_TONE — INTERPRETATION: Selective evidence work, not a
testing-basket allocation call.
MAIN_DRIVER — NEW ATTENTION / UNCHANGED FINANCIAL BASELINE: October 5
technical discussion reactivates July/August disclosures; TER's
September 29 product launch adds engineering capability, not bookings.
MARKET_CONTEXT — REPORTED_FACT: Revenue scopes and fiscal periods
differ. Aggregate demand language cannot determine each supplier's
conversion or retained margin.
ATTENTION_ENVIRONMENT — INTERPRETATION: Strong-cycle optimism meets a
warning about sentiment; current positioning and liquidity are
unmeasured.
WKAP_VIEW — UNRESOLVED_QUESTION: Does attributable business progress
justify the prevailing valuation? Quote observations below do not
answer that question.

## STATE_CHANGE_LEDGER

FORM — Previous evidence state: June-quarter HBM-linked DRAM growth
and a prospective Q3 guide. Current: same issuer evidence, newly
separated into normalized company execution and HBM attribution; first
research priority. No fresh financial promotion. Next changer:
matching-definition guide delivery and separately updated HBM
contribution.

TER — Previous: scaled memory-test business, lower next-quarter
revenue guide, new Magnum E2 platform. Current: unchanged financial
baseline with an explicit midpoint/margin test; second priority. A
product launch is not a booked-order transition. Next changer: actual
Q3 revenue/margin, with independent memory/HBM mix attribution.

AEHR — Previous: dated processor orders/backlog and a longer-term HBM
opportunity. Current: processor conversion remains trackable; the HBM
label cannot absorb the $22 million order. Third priority. This
refines classification, not issuer results or the historical
first-added date. Next changer: shipment/revenue/backlog
reconciliation; HBM qualification, production orders and revenue each
get their own dated state.

## CROSS_OBJECT_EVIDENCE_MATRIX

Use two status columns: COMPANY_EXECUTION and HBM_CONVICTION. Add
CONVERSION_STAGE so qualification cannot silently become revenue.
Initial states: FORM=reported profitable growth / HBM contribution
disclosed; TER=scaled revenue, lower forward guide / HBM included but
dollars unseparated; AEHR=contracted processor ramp, conversion
pending / longer-term opportunity without disclosed commercial HBM
proof.

FORM denominator worksheet — REPORTED_FACT: $84.999m − $57.057m =
$27.942m DRAM revenue increase. INTERPRETATION / CALCULATION:
approximately 75% × $27.942m ≈ $20.96m HBM contribution to that
increase. This is neither absolute HBM revenue nor a 75% growth rate.
The residual cannot be assigned to one technology without disclosure.

FORM margin worksheet — GAAP gross margin: 38.4% → 50.7%, +12.3
points; adjusted: 49.0% → 53.3%, +4.3 points. Restructuring
adjustment: 9.5 → 1.7 points, a 7.8-point reduction. That explains
much, not every decimal, of the difference between the two margin
changes. Do not attribute the entire GAAP rebound to HBM pricing. Q2
CFO $61.803m less capex $9.599m = $52.204m simple cash surplus, not
the issuer's differently adjusted free-cash-flow measure. Source:
https://investors.formfactor.com/news-releases/news-release-details/formfactor-inc-reports-2026-second-quarter-results

TER expectation worksheet — Q2 revenue $1,328.990m; Q3 guide
$1,200–1,300m. Midpoint $1,250m / $1,328.990m − 1 ≈ −5.94%. Even the
upper endpoint permits sequential decline. Memory above $200m for
three quarters establishes a level, not three growth observations or
$200m HBM revenue. Q2 GAAP operating margin fell 36.9% → 32.9%
sequentially. Our joint $1.25bn/32.9% execution test is analytical,
not a company margin forecast. Source:
https://investors.teradyne.com/news-events/press-releases/detail/445/teradyne-reports-second-quarter-2026-results

AEHR order worksheet — Keep May 29 backlog $80.6m, July 14 effective
backlog $100.6m and August 12 processor order $22m as dated
observations. Required roll-forward: opening comparable backlog + new
nonoverlapping bookings − order value removed under the issuer's
stated fulfillment policy − cancellations ± disclosed adjustments =
closing comparable backlog. Obtain that removal policy before
calculating; do not assume revenue recognition removes an order. Keep
shipments, shipped-but-not-recognized value and recognized revenue in
separate columns, without subtracting the same fulfillment twice.
Unknown removal or overlap data prevents a reconciled October balance.
FY27 $130–150m revenue guidance compares with FY26 approximately $50m,
but the fiscal-calendar change requires period labels. FY27 ends June
25, 2027. Sources:
https://www.aehr.com/2026/07/aehr-test-systems-reports-fiscal-2026-fourth-quarter-and-full-year-financial-results-with-record-quarterly-bookings-and-100-million-effective-backlog/
; https://www.sec.gov/Archives/edgar/data/1040470/000165495426006919/aehr_10k.htm

Funding control — AEHR's FY26 operating cash use of $3.3m and $97.4m
net stock-offering proceeds describe financing-supported liquidity,
not self-funded execution. Across objects, do not compare
different-period cash totals as a cash-quality league table. Separate
reported cash, capital spending and external financing before drawing
a return conclusion.

## RANKED_UNANSWERED_QUESTIONS

1. FORM: how much next-period growth remains specifically attributable
to HBM after normalizing restructuring and separating Triton? Retrieve
the next filing's segment explanation and matching GAAP/non-GAAP
reconciliation. This can constrain the leader even if total revenue
meets guidance.
2. TER: can memory mix support profitability through the planned
revenue step-down? Retrieve actual revenue, GAAP operating margin and
product commentary; distinguish HBM from NAND where disclosed. A
strong non-memory quarter can pass execution while leaving HBM
unchanged.
3. AEHR: which part of the processor order became shipment, recognized
revenue and cash, and what remains in a nonoverlapping backlog?
Retrieve the next results/filing and order updates. Customer identity
is not disclosed in the order and must not be inferred. HBM commercial
evidence is a separate request, not a prerequisite for recognizing a
successful processor business.

## SCENARIO_BRANCHES

Confirm: FORM delivers within its revenue guide with normalized
margins and continued attributable demand; TER meets the analytical
revenue/margin test; AEHR converts processor orders. Upgrade each
company-execution cell only as supported. HBM cells advance only with
specific attribution, and valuation remains outside this evidence
ranking.

Mixed: FORM meets total guidance through Triton; TER passes through
other testing categories; AEHR's processor ramp proceeds with no HBM
update. Company execution improves, HBM conviction remains
unchanged/unresolved. This is not an internally contradictory result.

Break: FORM revenue below $260m and adjusted gross margin below 52.5%,
without a quantified timing bridge, removes its execution lead. TER
below its $1.2bn guide floor with further GAAP operating-margin
deterioration is independently negative. AEHR material customer
pushouts or a ramp-related outlook cut invalidate timely processor
conversion. Silence alone triggers none of these failures.

Reorder: TER can move ahead when it passes while FORM fails. AEHR can
move above a failing TER if disclosed profitable processor conversion,
reconciled backlog and maintained outlook close its evidence gap; no
invented HBM sale is required. Relative promotion does not prove
superior expected equity returns.

NO_NEW_DATA: retain the last verified period and both verdicts. Record
the unanswered request, retrieval time and next future review date; do
not manufacture a financial state transition from another post.

## ATTENTION_TRADE_BOARD

Attention Trade Board

WKAP Attention View

Strongest already-reported HBM-growth link: FORM. Cleanest conceptual
asymmetry: AEHR's genuine processor opportunity can survive removing
assumed HBM revenue. Largest optionality/evidence gap and highest
qualitative fade risk: AEHR's HBM framing. Most crowded: unknown, not
inferable from three posts. Best durable-thesis candidate in this
evidence queue: FORM, conditional on normalized delivery and
valuation, not a proven stock-return winner.

## RADAR_OBJECT_INDEX

THESIS_OBJECT_1: FORM
THEME: AI testing / probe-card attribution and normalized margin
STATUS: active_trackable — rank 1, persistence diligence
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $148.77 USD (Nasdaq, U.S. premarket, October 5,
08:40 EDT /19:40 ICT; rechecked 19:43:38 ICT;
https://www.nasdaq.com/market-activity/stocks/form)
DATE_FIRST_ADDED_TO_RADAR: 2026-10-05
SETUP_TYPE: Realized growth → attribution persistence
ATTENTION_STAGE: Technical thesis reactivation
ATTENTION_WINDOW: 7–30 days; next relevant disclosure, not a promised
earnings date
KEY_QUESTION: What repeats after removing restructuring and separating Triton?

THESIS_OBJECT_2: TER
THEME: AI testing / memory mix and earnings conversion
STATUS: active_trackable — rank 2, guide/margin test
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $444.18 USD (Nasdaq, U.S. premarket, October 5,
08:41 EDT /19:41 ICT; rechecked 19:43:38 ICT;
https://www.nasdaq.com/market-activity/stocks/ter)
DATE_FIRST_ADDED_TO_RADAR: 2026-10-05
SETUP_TYPE: Scaled business → sequential mix test
ATTENTION_STAGE: Sector optimism with behavioral caution
ATTENTION_WINDOW: 7–30 days; product commercialization may take longer
KEY_QUESTION: Can actual profitability meet the joint test through
lower guided revenue?

THESIS_OBJECT_3: AEHR
THEME: AI testing / processor burn-in conversion and separate HBM opportunity
STATUS: active_trackable — rank 3, conversion watch
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $107.00 USD (Nasdaq, U.S. premarket, October 5,
08:43 EDT /19:43 ICT; rechecked 19:43:38 ICT;
https://www.nasdaq.com/market-activity/stocks/aehr)
DATE_FIRST_ADDED_TO_RADAR: 2026-07-09
SETUP_TYPE: Contracted orders → recognized economics
ATTENTION_STAGE: Category correction and ramp interpretation
ATTENTION_WINDOW: 7–30 days for updates; August shipment expectation
spans six months
KEY_QUESTION: What converts without counting the order twice or calling it HBM?

First-added audit covered all 70 public feeds through October 2. FORM
and TER had no earlier thesis cards; TER's earlier mentions were not
objects. AEHR's first verified card is July 9:
https://wkap.ai/radar/wkap-radar-feed-2026-07-09.html . Later reset
dates do not replace that history. Quotes are timestamped
observations, not synchronized executable prices or valuation
conclusions.

## THESIS OBJECTS

### THESIS_OBJECT_1 — FORM

CARD_ID: WKAP-RADAR-2026-10-05-FORM
CARD_TITLE: Attribute the Growth, Normalize the Margin
TYPE: trackable_wow
THEME: Probe-card test intensity / HBM attribution
STATUS: active_trackable — first research priority
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $148.77 USD (Nasdaq, U.S. premarket, October 5,
08:40 EDT /19:40 ICT; rechecked 19:43:38 ICT;
https://www.nasdaq.com/market-activity/stocks/form)
DATE_FIRST_ADDED_TO_RADAR: 2026-10-05
ATTENTION_STAGE: Technical thesis reactivation
ATTENTION_WINDOW: 7–30 days; next issuer evidence controls status

WHY_A_READER_CARES_NOW: The same testing narrative can produce an
accurate revenue update and an exaggerated margin explanation.
PREVIOUS_STATE: Reported HBM contribution plus prospective guidance,
without a fresh October result.
CURRENT_STATE: Same facts, split into attribution persistence and
normalized execution tests.
DECISION_OR_MONITORING_PAYOFF: Prioritize the next mix/reconciliation
table before extrapolating the GAAP rebound.

THESIS_SUMMARY

FORM leads because HBM is tied to a realized increment, not because
every revenue dollar or margin point is HBM. Triton supplies another
growth engine, making total-company success insufficient to prove HBM
acceleration.

WKAP_ANGLE

The surface-level frame: testing intensity explains the entire earnings jump.
The alternative frame: attributable demand and restructuring
normalization answer different questions.
The key research question: which repeatable mix and operating
economics remain after that separation?

CORE_THESIS

REPORTED_FACT: DRAM growth and margin reconciliations are in the
matrix. MANAGEMENT_EXPECTATION: Q3 revenue $260–280m; GAAP gross
margin 52% ±1.5 points; adjusted 54% ±1.5. INTERPRETATION:
inside-range revenue and matching-definition margins validate
execution; fresh HBM attribution validates that separate theme.
UNRESOLVED_QUESTION: durability and price paid for the earnings
stream.

ATTENTION_TRADE_FRAME

Attention Source: Carrio, October 5;
https://x.com/carrioresearch/status/2106959668279218564
Why Today: An engineering thesis is connected to a concrete reported line.
Attention Stage: Technical interpretation → repeatability test.
Attention vs Evidence — Hard evidence: filing attribution, not the
analyst's growth model.
Primary sources: https://investors.formfactor.com/node/20681/html ;
https://investors.formfactor.com/news-releases/news-release-details/formfactor-inc-reports-2026-second-quarter-results
Attention / interpretation: DRAM above $80m is Carrio's checkpoint,
not management guidance; no holding disclosure visible.
Attention Path: Complexity → probe demand → attributable revenue →
normalized earnings.
Attention Asymmetry: Clearer growth attribution coexists with a less
dramatic adjusted-margin gain.
Crowding Risk: Unmeasured; repeated HBM labels can hide denominator errors.
What Could Sustain Attention: New attributable demand and retained
normalized margin.
What Could Make Attention Fade: Repeated intensity arguments without
incremental filings.
Attention-to-Thesis Conversion: Persistence supports execution and
attribution independently, then valuation work tests allocation.

WEAKEST_ASSUMPTION: The incremental mix and normalized profitability
repeat rather than being a temporary customer/product configuration.
MOST_IMPORTANT_DATA_POINT: Next DRAM/HBM attribution paired with the
matching gross-margin reconciliation.
NEXT_DATA_POINT: Next issuer results/filing, date not established
here. Revenue below $260m plus adjusted margin below 52.5%, absent a
quantified timing bridge, reverses execution promotion. Missing HBM
detail leaves attribution unresolved, not failed.

### THESIS_OBJECT_2 — TER

CARD_ID: WKAP-RADAR-2026-10-05-TER
CARD_TITLE: Judge the Guide Before the Cycle Story
TYPE: trackable_wow
THEME: Memory test / mix and operating conversion
STATUS: active_trackable — second research priority
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $444.18 USD (Nasdaq, U.S. premarket, October 5,
08:41 EDT /19:41 ICT; rechecked 19:43:38 ICT;
https://www.nasdaq.com/market-activity/stocks/ter)
DATE_FIRST_ADDED_TO_RADAR: 2026-10-05
ATTENTION_STAGE: Sector optimism with behavioral caution
ATTENTION_WINDOW: 7–30 days; launch-to-commercialization timing unresolved

WHY_A_READER_CARES_NOW: Calling a guided sequential decline a surprise
miss can damage the model as much as ignoring a genuine margin
deterioration.
PREVIOUS_STATE: Large memory business and a lower prospective revenue range.
CURRENT_STATE: Same baseline, explicit joint revenue/margin hurdle,
separate HBM attribution cell.
DECISION_OR_MONITORING_PAYOFF: Evaluate expectation delivery before
promoting the broad testing narrative.

THESIS_SUMMARY

Established scale earns second place; neither a memory revenue floor
nor a product announcement establishes accelerating HBM earnings.
Preserve the difference between total-company execution and
device-specific demand.

WKAP_ANGLE

The surface-level frame: a strong packaging/testing cycle means
sequential growth.
The alternative frame: the company's own guide permits contraction
while mix determines profitability.
The key research question: can actual revenue and operating margin
jointly meet the defined test?

CORE_THESIS

REPORTED_FACT: Q2 revenue and operating margin establish the starting
point. MANAGEMENT_EXPECTATION: Q3 $1.2–1.3bn. INTERPRETATION: revenue
at least $1.25bn and GAAP operating margin at least 32.9% constitute
our execution-promotion test, not issuer guidance.
UNRESOLVED_QUESTION: incremental HBM contribution and Magnum E2
commercial timing.

ATTENTION_TRADE_FRAME

Attention Source: Herman Jin, October 5;
https://x.com/ShanghaoJin/status/2107061094967833081
Why Today: It counters the inference that strong industry fundamentals
eliminate sentiment risk.
Attention Stage: Sector thesis plus behavioral caution, not company results.
Attention vs Evidence — Hard evidence: issuer revenue, guide and
product specifications.
Primary sources:
https://investors.teradyne.com/news-events/press-releases/detail/445/teradyne-reports-second-quarter-2026-results
; https://investors.teradyne.com/news-events/press-releases/detail/451/teradyne-announces-magnum-e2-to-address-next-generation-memory-test-demands-driven-by-ai-and-high-performance-computing
Attention / interpretation: Optimistic cycle outlook and
liquidity/sentiment warning are opinion, not measured flow; no holding
disclosure visible.
Attention Path: Cycle narrative → customer demand → reported mix →
operating profit.
Attention Asymmetry: Commercial scale can be real while the
next-quarter hurdle is lower.
Crowding Risk: Unmeasured; broad-cycle extrapolation is the
identifiable reasoning risk.
What Could Sustain Attention: Commercial detail and financial
delivery, not specifications alone.
What Could Make Attention Fade: Repeated launch coverage without
conversion evidence.
Attention-to-Thesis Conversion: Joint execution pass; HBM upgrade
additionally requires attributed HBM evidence.

WEAKEST_ASSUMPTION: Memory mix preserves profitability through the
guided step-down.
MOST_IMPORTANT_DATA_POINT: Actual Q3 revenue with GAAP operating
margin and disaggregated memory commentary.
NEXT_DATA_POINT: Next financial disclosure; Magnum E2
commercialization is a separate track. Below $1.2bn with margin below
32.9% demotes execution; an in-range decline alone does not.

### THESIS_OBJECT_3 — AEHR

CARD_ID: WKAP-RADAR-2026-10-05-AEHR
CARD_TITLE: Convert the Processor Order; Keep HBM Separate
TYPE: trackable_wow
THEME: Reliability burn-in / contracted ramp
STATUS: active_trackable — third research priority
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $107.00 USD (Nasdaq, U.S. premarket, October 5,
08:43 EDT /19:43 ICT; rechecked 19:43:38 ICT;
https://www.nasdaq.com/market-activity/stocks/aehr)
DATE_FIRST_ADDED_TO_RADAR: 2026-07-09
ATTENTION_STAGE: Category correction and ramp interpretation
ATTENTION_WINDOW: 7–30 days for updates; six-month shipment
expectation is longer

WHY_A_READER_CARES_NOW: Removing assumed HBM revenue prevents a
modeling error without erasing an actual processor business.
PREVIOUS_STATE: Dated processor bookings/backlog and long-term HBM opportunity.
CURRENT_STATE: Processor conversion watch retained; no automatic
commercial HBM promotion.
DECISION_OR_MONITORING_PAYOFF: Build a nonoverlapping order-to-revenue
bridge instead of adding headline backlog figures.

THESIS_SUMMARY

The size of the contracted ramp makes conversion consequential
relative to the old revenue base. The investment uncertainty is
execution and economics, not whether a processor burn-in order can be
renamed an HBM order.

WKAP_ANGLE

The surface-level frame: protecting costly HBM packaging makes Aehr an
established HBM-device testing supplier.
The alternative frame: the economic rationale for screening processors
does not change the tested device.
The key research question: what was delivered, recognized and
collected, with what remaining obligation?

CORE_THESIS

REPORTED_FACT: August's $22m processor order and July's dated backlog.
MANAGEMENT_EXPECTATION: six-month shipment plan and FY27 $130–150m
revenue guide. INTERPRETATION: timely profitable conversion can
promote company research priority independently of HBM.
UNRESOLVED_QUESTION: current reconciled backlog and separately
commercial HBM evidence. Qualification, a production order and
recognized revenue are distinct stages, not equivalent proof.

ATTENTION_TRADE_FRAME

Attention Source: Carrio, October 5;
https://x.com/carrioresearch/status/2106966041561899075
Why Today: It distinguishes burn-in from probing/testing and current
programs from HBM optionality.
Attention Stage: Technical classification → order-conversion diligence.
Attention vs Evidence — Hard evidence: disclosed processor order,
dated results and 10-K.
Primary sources:
https://www.aehr.com/2026/08/aehr-receives-22-million-follow-on-order-for-ai-processor-wafer-level-burn-in-systems/
; https://www.sec.gov/Archives/edgar/data/1040470/000165495426006919/aehr_10k.htm
Attention / interpretation: Same author as FORM, not independent
corroboration; no holding disclosure visible. His backlog framing is
dated against July 14, not accepted as October's current balance.
Attention Path: Contract → shipment → revenue → profitable conversion;
HBM has its own qualification/order/revenue path.
Attention Asymmetry: Processor progress can matter even with no change
to HBM optionality.
Crowding Risk: Unmeasured; largest risk here is thematic relabeling.
What Could Sustain Attention: Reconciled conversion, maintained
outlook and explicit commercial milestones.
What Could Make Attention Fade: Old backlog repeatedly presented as
current or HBM plans substituted for orders.
Attention-to-Thesis Conversion: Processor economics close the
company-evidence gap; only attributed HBM milestones advance HBM
status.

WEAKEST_ASSUMPTION: Customers accept the contracted ramp on the
expected timing with profitable economics.
MOST_IMPORTANT_DATA_POINT: Shipment/recognized-revenue update and a
comparable backlog roll-forward.
NEXT_DATA_POINT: Next order/results/filing update; no new quarterly
date established. Material pushouts, cancellations or a ramp-related
guide cut invalidate timely conversion, not necessarily all long-term
opportunity.

## CROSS_OBJECT_ATTENTION_COMPARISON

Cross-Object Attention Comparison

Cleanest Attention Trade: AEHR's category correction as a research
setup, not an entry call. Most Evidence-Backed Attention Trade: FORM.
Most Crowded Attention Trade: unknown. Highest Fade Risk: unsupported
AEHR HBM framing. Best Candidate to Become a Durable Thesis: FORM
conditional on persistence and valuation. Relative research rank and
absolute thesis status must remain separate.

## 7_DAY_RESEARCH_WORKFLOW

Review by October 12, 2026, or earlier upon a relevant disclosure. Use
current issuer indexes, not only historical links. FORM:
https://investors.formfactor.com/press-releases ; TER:
https://investors.teradyne.com/news-events/press-releases ; AEHR:
https://www.aehr.com/press-releases/ . Follow new dated financial
disclosures to their exact release/filing and retain the publication
time. These are research deadlines, not invented reporting dates.

### FORM — 7-Day Checks

- Retrieve the latest mix and margin reconciliation; keep GAAP and
adjusted measures in parallel columns.
- Recalculate the DRAM increment and HBM attribution only if a new
denominator is disclosed. Do not backsolve total HBM revenue.
- Separate Triton/Systems from DRAM and ask which engine explains
company execution. Preserve the two verdicts if attribution is absent.

### TER — 7-Day Checks

- Compare any result with the original $1.2–1.3bn range before calling
it a beat/miss; test our midpoint/margin pair separately.
- Extract memory scope and HBM/NAND qualifiers. Do not turn three
quarters above a level into a growth streak.
- Search Magnum E2 updates for customer/order/shipment evidence, not
repeated specifications; preserve product launch as a nonfinancial
milestone.

### AEHR — 7-Day Checks

- Build the dated backlog roll-forward using the issuer's removal
policy, with overlap/cancellation adjustments. Track shipments,
shipped-but-not-recognized value and revenue separately; leave missing
inputs unresolved rather than force equality.
- Tag every order by tested device, customer disclosure and expected
timing; never infer the unnamed processor customer's identity.
- Keep HBM qualification, production order and revenue fields
separate. Check the calendar's year: the visible October 6 results
item belongs to 2025, not tomorrow.

### Cross-Object — 7-Day Checks

Return one row per changed assumption: object | source/date |
period/unit/denominator | evidence type | previous → current |
conversion stage | company-execution effect | HBM effect | rank effect
| missing input | next review. Test the adjacent-pair reversals in
SCENARIO_BRANCHES. No change is a valid output. If NO_NEW_DATA, retain
both states and set the next review seven calendar days after the
actual ICT review date, or an earlier confirmed future event; never
reuse an already elapsed date.

Primary-source retrieval registry: FORM:
https://investors.formfactor.com/press-releases ; TER:
https://investors.teradyne.com/news-events/press-releases ; AEHR:
https://www.aehr.com/press-releases/ ; exact financial baselines are
in the matrix/cards.

Attention-source retrieval registry: FORM:
https://x.com/carrioresearch/status/2106959668279218564 ; TER:
https://x.com/ShanghaoJin/status/2107061094967833081 ; AEHR:
https://x.com/carrioresearch/status/2106966041561899075

Evidence-state snapshot: FORM=attributed growth/persistence open;
TER=scaled memory/lower guide/HBM dollars unseparated; AEHR=processor
ramp pending/HBM optionality; classification=trackable_wow;
update-rule=primary evidence changes the relevant verdict, attention
changes the question; unknowns remain unknown.

## 30_DAY_RESEARCH_WORKFLOW

Review by November 4, 2026 using the same forward routes, or earlier
when relevant evidence appears. October 13 conference materials may
improve detail but are not promised results or guaranteed public
financial updates. Retain source dates and fiscal calendars. If
required observations have not occurred, return NO_NEW_DATA and
schedule a genuinely future review under the 7-day rule, not a
fabricated verdict.

### FORM — 30-Day Checks

Validate: guide delivery with matching-definition margins. Upgrade:
persistent attributed growth and normalized profitability; allocation
still requires valuation. Downgrade: weaker underlying economics
despite a cosmetic GAAP rebound. Invalidate the execution lead: joint
below-floor revenue/adjusted-margin failure without quantified timing
repair. An execution pass through Triton alone does not upgrade HBM
conviction.

### TER — 30-Day Checks

Validate: revenue within the guide and explained mix. Upgrade
execution: at least $1.25bn revenue and 32.9% GAAP operating margin;
strengthen HBM only with attribution. Downgrade: weaker conversion or
profitability. Invalidate the near-term execution case: below-range
revenue with further margin deterioration. Specifications without
sales remain product evidence, not financial confirmation.

### AEHR — 30-Day Checks

Validate: disclosed shipments and recognized revenue consistent with
the contracted ramp, reconciled backlog and maintained guide. Upgrade:
profitable processor conversion closes the evidence gap, even if HBM
is unchanged. Downgrade: timing or economics worsen. Invalidate timely
conversion: material pushout/cancellation or ramp-related outlook cut.
Separately advance HBM only to the specific qualification, order or
revenue stage actually disclosed.

### Cross-Object — 30-Day Checks

Rebuild FORM > TER > AEHR rather than inherit it. A failing FORM and
passing TER can reverse the first pair; verified AEHR processor
conversion can outrank a failing TER. Neither outcome alone
establishes a purchase opportunity. Before allocating, request
comparable valuation/downside cases using normalized earnings, funding
and dilution; this Feed deliberately does not fill missing valuation
inputs with thematic enthusiasm.

## WKAP DAILY TOP 3

Three market sources worth feeding into today’s market chat. Not
required reading — WKAP has already extracted the signal.

### 1. @carrioresearch — Test Intensity Needs the Right Denominator

URL: https://x.com/carrioresearch/status/2106959668279218564
WKAP signal: October 5 independent engineering/research interpretation
connects FORM to DRAM probe-card revenue; the author's $80m checkpoint
is opinion, not guidance. No holding disclosure visible.
Why it matters today: It turns an abstract testing thesis into a
filing question while exposing the difference between HBM's
contribution to an increment and total HBM sales.
Themes/tickers: Probe cards, HBM attribution, normalized margins / FORM.
Question to ask: Which disclosed denominator supports the growth
attribution, and what new mix/reconciliation evidence would show that
the economics repeat?

### 2. @ShanghaoJin — A Strong Cycle Does Not Remove the Earnings Hurdle

URL: https://x.com/ShanghaoJin/status/2107061094967833081
WKAP signal: October 5 sector-cycle optimism paired with
sentiment/liquidity caution; interpretation, not measured flow or TER
guidance. No holding disclosure visible.
Why it matters today: It challenges basket-level extrapolation while
TER's own guide permits a sequential step-down.
Themes/tickers: Packaging/testing, expectation discipline, memory mix / TER.
Question to ask: What actual revenue/margin and separately attributed
HBM evidence would justify changing company execution and thematic
conviction independently?

### 3. @carrioresearch — Processor Burn-In Is Not an HBM Order

URL: https://x.com/carrioresearch/status/2106966041561899075
WKAP signal: October 5 distinction among burn-in, testers and probe
cards, and between current programs and HBM potential; same author as
item 1, not independent corroboration. No holding disclosure visible.
Why it matters today: A real $22m processor order can improve the
company case without establishing commercial HBM-device revenue; the
cited backlog needs its July date.
Themes/tickers: Reliability testing, order conversion, HBM optionality / AEHR.
Question to ask: What nonoverlapping shipment/revenue/backlog bridge
validates the processor ramp, and which separately identified HBM
milestone would actually change the HBM state?

WKAP Radar Feed - 2026 - 10 - 05

Agent-readable facts