Preliminary brief— based on press release only. Full analysis including management tone and Q&A will be added when the transcript is available.

WDC · Q3 2026 Earnings

Western Digital

Reported April 30, 2026

30-second summary

Western Digital crushed its own Q3 guide on every line — revenue $3.34B (vs. $3.1–3.3B), non-GAAP EPS $2.72 (vs. $2.15–2.45), gross margin 50.5% (vs. 47–48%) — and the Q4 guide pushes harder still: revenue $3.55–3.75B implying 36–44% YoY and gross margin stepping to 51–52%. The 38% margin algorithm flagged three quarters ago is now ~1,250bps in the rearview; agentic AI is being framed as a "step function" demand event with LTAs extending to 2029.

Headline numbers

EPS

Q3 FY2026

$2.72

Revenue

Q3 FY2026

$3.34B

+45.0% YoY

Gross margin

Q3 FY2026

50.5%

Free cash flow

Q3 FY2026

$0.98B

Operating margin

Q3 FY2026

38.6%

Key financials

Q3 FY2026

Metric	Q3 FY2026	YoY	Q2 FY2026	QoQ
Revenue	$3.34B	+45.0%	$3.02B	+10.6%
EPS	$2.72	—	$2.13	+27.7%
Gross margin	50.5%	—	46.1%	+440bps
Operating margin	38.6%	—	33.8%	+480bps
Free cash flow	$0.98B	—	$0.65B	+49.8%

Guidance

Western Digital posted strong beats across revenue, EPS, and margins in Q3 FY2026, with Q4 FY2026 guidance implying accelerated growth and margin expansion despite significant revenue upside already achieved.

Guidance is issued for both next quarter and the full year. Both may appear below.

Actuals vs prior guidance

Metric	Period	Prior guide	Actual	Δ	Result
Revenue	Q3 FY2026	$3.1B to $3.3B	$3.337B	+$0.037B above high end of guide	Beat
Revenue YoY growth	Q3 FY2026	approximately 40% YoY	+45%	+5 percentage points above guide	Beat
Non-GAAP EPS	Q3 FY2026	$2.15 to $2.45	$2.72	+$0.27 above high end of guide	Beat
Gross margin	Q3 FY2026	47% to 48%	50.5%	+2.5 to +3.5 percentage points above guide range	Beat
Operating expenses	Q3 FY2026	$380M to $390M	not disclosed	in-line (actual not reported separately)	Met

New guidance

Metric	Period	Guide	YoY
Revenue	Q4 FY2026	$3.55B to $3.75B	+36% to +44%
Non-GAAP EPS	Q4 FY2026	$3.1 to $3.4	—
Gross margin	Q4 FY2026	51% to 52%	—
Operating expenses	Q4 FY2026	$385M to $395M	—
Interest and other expense, net	Q4 FY2026	~$10M	—

Management tone

Customer optimization → structural demand reset → timelines pulled forward → customer lock-in to 2028 → agentic AI as compounding loop.

The demand framing has moved from "AI tailwind" to "compounding loop." Two quarters ago the story was inference workloads generating storage need. This quarter management collapsed three workload categories into a single feedback mechanism: "These forces are not additive. They are a compounding loop. Inference creates data. Agents consume and generate more data. Physical AI creates data and trains synthetic models that create more data. And ultimately, the loop accelerates." What changed is the implied derivative — a loop accelerates by definition; a tailwind merely persists. This is the strongest articulation of structural demand the company has offered post-spin.

The exabyte growth ceiling has been broken — language is now "step function," not "23% upside case." Q1 FY26 framed demand as trending toward 23% with potential upside in 2027–28. Q2 FY26 left the low-20s framing intact. This quarter management said "we expect agentic AI to drive a step function increase in capacity-oriented storage demand." "Step function" is a discrete-jump word, not a continuous-growth word. Combined with LTAs now extending to 2029, management is asserting that the historical 25% CAGR is the floor of a higher regime, not the ceiling of the current one.

LTA horizon extended again — from 2028 to 2029. Q4 FY25 had top-five hyperscalers under PO/LTA for FY26. Q1 FY26 added one hyperscaler for all of calendar 2027. Q2 FY26 extended to "two through 2027, one through 2028." This quarter: "we now have LTAs that extend into calendar year 29." The visibility window has lengthened by 12+ months for the fourth consecutive quarter. Pricing inside those LTAs is framed as moving to "high single-digit" increases later in 2026.

HAMR has moved from "one customer in qualification" to four. Q1 FY26 pulled HAMR qualification forward by six months to one customer in H1 2026. Q2 FY26 said qualification "started this month with one customer." This quarter: "we are now in qualification with 4 customers... our Hammer development is going really well." The dual-track ePMR+HAMR strategy is being executed at pace, and four customers in qualification is the high end of last quarter's "up to three by year-end" target — hit a quarter ahead of plan.

UltraSMR has scaled from "top three adopted" to a 60%-of-exabytes target. Q2 FY26 had top three customers fully adopted and 2–3 more in process. This quarter management put a number on the target: "close to about 60% of all the exabytes that we ship will be on UltraSMR... we plan to have all of our major customers qualified on UltraSMR by the end of calendar year 2027." A capacity-uplift technology going from selective adoption to industry-default within ~18 months is the cleanest version of the no-capex margin-accretion story.

Recurring themes management leaned on this quarter:

Agentic AI and physical AI driving structural step-function increase in storage demand beyond historical 25% CAGRCompounding data generation loop from inference, agents, synthetic data, and physical AI creating multi-year visibility into customer demandGross margin expansion trajectory (50.5% current, guidance 51-52%) driven by pricing, mix shift to higher-capacity drives, and cost disciplineLTA framework extending into 2028-2029 with exabyte-based pricing structure providing multi-year revenue visibility and predictabilityHAMR qualification acceleration ahead of schedule with dual-track strategy reducing customer transition risk while maintaining ePMR density improvementsUltraSMR adoption expanding from three largest customers to near-universal deployment target (60% of exabytes by FY27 end, all major customers by calendar 2027)

Risks management surfaced:

Hammer reliability and yields still require optimization before mainstream production transitionCustomer adoption curve for new products (40TB ePMR, Hammer) will phase in improvements rather than immediate impactVariability in LTA pricing beyond first year and potential for customer demand divergence by geography or business modelFlash pricing parity risk in large-scale object storage workloads if NAND cost-per-byte declines accelerateSupply chain efficiency and procurement cost pressures as exabyte volumes scale significantly

Answers to last quarter's watch list

Whether Q3 revenue lands above the $3.3B high end. Yes — $3.34B printed, $37M above the high end, materially exceeding the ~$15–20M beat pattern of the prior three quarters.

Resolved positively

Whether gross margin reaches 48%. Decisively exceeded. 50.5% printed against a 47–48% guide — a +250bps beat above the high end, the largest in the post-spin run. Q4 guide is 51–52%.

Resolved positively

Innovation Day disclosures (Feb 3). Not specifically called out in the press release. The exabyte framing has moved to "step function" language and the 38% margin algorithm is now ~1,250bps below the current print, but a formal CAGR reset or replacement margin anchor was not disclosed.

Continue monitoring

SanDisk share monetization timing. The press release did not disclose execution detail on the 7.5M-share monetization.

Continue monitoring

Whether a fourth customer signs an LTA extending into 2027 or 2028. Management said LTAs now extend into calendar year 2029. The implication is that the 2027/2028 horizon flagged last quarter has been pushed further out, though the customer count behind the 2029 extension wasn't quantified in the press release.

Resolved positively

Capital return cadence. The press release disclosed only the $0.15 quarterly dividend; no buyback execution figures or repurchase pace was called out. The Q2 $1.4B return (>100% of FCF) cannot be confirmed as continuing.

Continue monitoring

What to watch into next quarter

Whether Q4 revenue lands above the $3.75B high end. The pattern of beating the top of the range widened to $37M this quarter from ~$15–20M prior; if that holds or expands, Q4 prints closer to $3.8B and YoY exceeds +44%.

Whether gross margin reaches 52%. Q4 guide midpoint is 51.5%; Q3 cleared its high end by 250bps. A 53%+ print would extend the margin slope past where any prior framework anchored.

The customer count behind the 2029 LTAs. Management said LTAs extend into 2029 but did not quantify how many customers. A second or third 2029 commitment would lock in supply visibility past most peer industry frameworks.

HAMR ramp transition pace. Four customers are in qualification; watch for the first commercial revenue contribution and whether the ePMR–HAMR mix shift starts showing up in gross margin attribution.

Pricing realization in LTAs. Management flagged high-single-digit pricing increases moving into late calendar 2026. Watch whether Q4 ASP/TB commentary confirms this is showing up in the print rather than only in forward commitments.

Whether net interest expense actually steps to ~$10M. The Q4 guide implies a $40M step-down from Q3's ~$50M run-rate. Execution on the SanDisk monetization or debt paydown would explain it; absence of either would create an EPS-quality question.

First quantified FY27 framing. Four consecutive quarters without an FY revenue range. With LTAs now to 2029 and demand framed as "step function," the absence of an FY framework becomes harder to justify each quarter.

Sources

Western Digital Q3 FY2026 press release, filed 2026-04-30 — https://www.sec.gov/Archives/edgar/data/106040/000162828026028878/a4ex991-pressreleaseq326.htm
Western Digital Q3 FY2026 prepared remarks (as extracted; no full Q&A transcript available)
Western Digital Q1 and Q2 FY2026 and Q4 FY2025 prior tapebrief coverage (for trend comparisons)

Get the next brief, free.

We publish analyst-grade earnings briefs the same day or morning after every call — headline numbers, segment KPIs, Q&A highlights, and tone analysis. Free during beta.

This is not investment advice.