Mobileye's 69% Customer Anchor Just Got a 2029 Expiration Date insight cover
Industry NewsMBLY · NVDA · QCOM•14 min read

Mobileye's 69% Customer Anchor Just Got a 2029 Expiration Date

Hyundai Motor Group told Reuters on Sep 12 it will push its in-house Level 2 driver-assist stack to late 2029 and use NVIDIA DRIVE Hyperion as the 2028 bridge. For Mobileye, that one customer accounted for 69% of revenue in its Q2 2026 10-Q — and it is now openly plotting to stop buying. The $24.5B pipeline narrative absorbs the headline hit but can't outrun the math: NVIDIA and Qualcomm are the ones collecting the 2028 revenue, Hyundai Mobis absorbs the integration work, and TSMC keeps the foundry order either way.

Published Sep 13, 2026Updated Sep 13, 2026

Mobileye Q2 2026 revenue

$508M

Flat YoY; FY2026 guide midpoint $1.995B (4–7% growth), per Jul 23, 2026 release

One-customer revenue share

69%

Per Mobileye 10-Q for the six months ended Jun 27, 2026 (Note 9)

Mobileye 8-year pipeline

$24.5B

As of YE2025, +42% from prior update per Jan 22, 2026 release

MBLY market cap / EV/Sales TTM

$6.7B / 2.7x

Stock $8.25 vs. 52-week high $15.81, as of Sep 13, 2026

Hyundai Motor Group president Park Min-woo tried to soften the news, telling reporters on Sep 12 that Hyundai's NVIDIA partnership is \"not about leaving our destiny entirely in their hands.\" The number that mattered, though, sat in a footnote: Mobileye's own Q2 2026 10-Q discloses that one major customer generated 69% of total revenue in the six months ended June 27, 2026. That filing is dated July 23, 2026 — three weeks before Reuters confirmed the customer is preparing to replace EyeQ with its own stack. A 2028 NVIDIA bridge plus a 2029 in-house launch is not a delay story. It is an exit roadmap.

The 69% Anchor — Why Mobileye's Customer Math Just Tilted

Mobileye's six months ended June 27, 2026 generated roughly $1.07B in revenue. The 10-Q footnote (Note 9) flags a single customer at 69% of the period total — i.e., ~$735M of revenue over two quarters tied to one automaker. Cross-reference the timing: Hyundai Motor Group retailed 999,015 vehicles globally in Q2 2026 and 1,948,377 in H1 2026, and Hyundai Mobis sits as the in-house Tier-1 wiring EyeQ into nearly every Hyundai and Kia ADAS-equipped model. The 69% is, in practical terms, Hyundai Group exposure.

On a $1.97–2.02B FY2026 revenue guide (Mobileye Q2 release, Jul 23, 2026), the implied exposure is roughly $1.36–1.39B from a single OEM. If late-2029 is the inflection — when Hyundai says its own Level 2 and Level 2+ software reaches production — Mobileye's loss of one mid-decade refresh cycle from its anchor customer sets the floor on the revenue cliff the Street has to model.

Mobileye Q2 2026 revenue

$508M

Flat YoY; FY2026 guide midpoint $1.995B (4–7% growth), per Jul 23, 2026 release

One-customer revenue share

69%

Per Mobileye 10-Q for the six months ended Jun 27, 2026 (Note 9)

Mobileye 8-year pipeline

$24.5B

As of YE2025, +42% from prior update per Jan 22, 2026 release

MBLY market cap / EV/Sales TTM

$6.7B / 2.7x

Stock $8.25 vs. 52-week high $15.81, as of Sep 13, 2026

A single-customer disclosure above the SEC's 10% threshold is not unusual. A single customer at 69% of revenue, publicly plotting to bring the work in-house by late 2029, is the kind of concentration that forces a re-rating before it shows up in the income statement.

The NVIDIA Bridge — Real Money, Mostly Priced

The 2028 Hyundai-NVIDIA deployment rides on the DRIVE Hyperion architecture formalized in the March 16, 2026 Hyundai Motor / Kia / NVIDIA expansion. NVIDIA's Automotive segment printed $2.35B in FY2026 (Jan year-end), up 38.7% YoY, with Q2 FY2026 alone at $586M (+69% YoY). Add the $96.2B total Q2 FY2027 revenue reported Aug 26, 2026, and NVIDIA has the cash-flow base to fund a Hyundai-grade platform deal without strain. The question is whether the Street has over-learned this lesson.

NVIDIA Automotive segment revenue growth

FY ends late January. FY2026 includes the $586M Q2 print that anchored the Hyundai DRIVE Hyperion expansion.

Unit: USD billions

FY2024

Per NVIDIA FY2024 10-K

1.1

FY2025

+57% YoY

1.7

FY2026

+38.7% YoY

2.4

On January 7, 2026, Hyundai Motor shares jumped as much as 14.9% to a record high on speculation of a deeper NVIDIA tie-up — that was before the March 16 confirmation and the Sep 12 in-house roadmap disclosure. The remaining alpha on NVIDIA's auto win is now thinner: investors who missed the January rally are paying Sep-2026 prices for an already-codified deal, and NVIDIA's own auto mix is still under 1% of revenue. NVIDIA's stock trades at 17.5x EV/Sales TTM versus Mobileye's 2.7x — the multiple expansion is the easy part; the absolute auto-revenue contribution stays rounding-error small even after the Hyundai win.

Qualcomm's Snapdragon Ride Flex — The Counter-Bid That Actually Crossed

The cleanest counter to a Mobileye-or-NVIDIA narrative is the Hyundai Mobis + Qualcomm comprehensive agreement signed at CES 2026 (Jan 7, 2026). Mobis is integrating the Snapdragon Ride Flex SoC with Mobis's own unified software stack — meaning even if Hyundai eventually runs proprietary AI on top, the silicon inside could still be Qualcomm. Qualcomm disclosed in Q2 FY2026 (Apr 29, 2026) that automotive revenue exceeded $5B annualized for the first time, with management guiding to $4B in car revenue in 2026 and a $15B ADAS pipeline. The competitive question is whether Snapdragon Ride Flex is the chip inside Hyundai's 2028 NVIDIA-based Level 2 systems or whether it gets displaced when Hyundai's in-house stack takes over in late 2029.

ADAS silicon contenders for the Hyundai Group wallet, 2026–2029
VendorRole in HMG roadmapAuto-revenue scalePublic source
MobileyeIncumbent; 69% concentration risk; 2029 in-house replacement signaled$508M Q2 2026 revenue; FY2026 guide $1.97–2.02BMobileye 10-Q (Jun 27, 2026); Jul 23, 2026 release
NVIDIA2028 Level 2/2+ bridge via DRIVE Hyperion$2.35B FY2026 auto segment (+38.7% YoY)NVIDIA FY2026 10-K; Q2 FY2027 release Aug 26, 2026
QualcommSnapdragon Ride Flex with Hyundai Mobis (Jan 2026); could remain silicon inside in-house stack>$5B annualized Q2 FY2026; $15B ADAS pipelineQualcomm Q2 FY2026 release; CES 2026 joint announcement
TSMCFabs EyeQ (incumbent) and likely DRIVE Thor + Snapdragon Ride Flex (successors)Auto is one slice of multi-vertical foundry demandIndustry reporting; Mobileye disclosures
Qualcomm's $15B ADAS pipeline, signed partnerships with Hyundai Mobis, and the >$5B annualized automotive run-rate make it the only vendor with credible claims to the silicon slot inside Hyundai's in-house stack — a path Mobileye has no obvious route into.

The Tier-1 Power Shift — Hyundai Mobis Captures the Integration Layer

Every version of the Hyundai ADAS story routes through Hyundai Mobis. Mobis is the captive Tier-1 that has integrated EyeQ into Hyundai-Kia vehicles for years; it signed the Snapdragon Ride Flex agreement with Qualcomm on Jan 7, 2026; and it inherits the 2028 NVIDIA DRIVE Hyperion integration work by virtue of being the platform lead for Hyundai Motor Group. Q2 2026 revenue of KRW 16.3T (+2.4% YoY) — disclosed in Mobis's Aug 14, 2026 earnings summary — captures the full ADAS bill of materials, regardless of whose silicon is inside.

This is where the thesis becomes counter-intuitive: the vertical-integration push is bearish for Mobileye but neutral-to-bullish for Mobis. Hyundai's vertical integration is, operationally, an extension of Mobis's existing mandate — and Mobis has already hedged by dual-qualifying Qualcomm silicon. The 2029 in-house software launch is, from Mobis's perspective, more software IP to monetize across the 5.55M-vehicle 2030 target Hyundai Group laid out at its August 2026 Investor Day, not less.

Pipeline Math vs. Customer Math — Why Both Can Be Right at Once

The bull case for Mobileye is the $24.5B eight-year forward automotive revenue pipeline that the company disclosed at year-end 2025, up 42% from the prior update. That pipeline is meant to absorb the Hyundai blow: more SuperVision wins, the Stellantis cloud-enhanced ADAS agreement signed Jul 21, 2026, the Volkswagen robotaxi expansion, the major U.S. OEM Surround ADAS win announced Jan 5, 2026 (with 19M+ system deliveries estimated). The case is structurally coherent — Mobileye is no longer dependent on any one OEM in theory.

The problem is timing. The Q2 2026 print was $508M — flat YoY despite +3% EyeQ volume growth — and FY2026 guidance is a modest 4–7%. The pipeline does not start offsetting Hyundai attrition until the post-2029 cohort of wins ramps, and those wins must out-compete NVIDIA DRIVE Hyperion and Qualcomm Snapdragon Ride Flex on cost per TOPS, not just feature parity. In the interim, Mobileye's stock at $8.25 is already 48% off its 52-week high — the market has begun pricing the gap between pipeline potential and customer concentration.

  • EyeQ SoCs ran ~91% of Mobileye revenue in Q2 2026 (10-Q and Q2 release), so the Hyundai attrition hits the highest-margin product line directly.
  • Mobileye shipped ~10M EyeQ units in Q2 2026 (+3% YoY), outperforming its top-10 customers' production by 8 points — volume is fine; customer mix is the problem.
  • The Jul 23, 2026 release also flagged a CEO transition and a leadership reshuffle — an awkward backdrop for a strategic pivot that needs stable customer-facing execution.
  • Intel still owns ~96.9% of Mobileye voting power (50M Class A + all Class B shares per 2026 proxy), so any capital allocation response must clear Intel's board first.

Two Horizons, Two Verdicts

Short-term (days to quarters): The Sep 12 disclosure is a thesis-changing fact but not yet a print-changing one. Hyundai's 2028 Level 2 vehicles will still contain EyeQ silicon in many of the older-platform variants alongside DRIVE Hyperion units, and Mobileye's FY2026 guide has room to land even with Hyundai churn baked in. The Q3 2026 print (expected late October) will be the first datapoint where the buy-side can test whether major-customer share has begun to erode. Watch for two things: any change to the one-customer disclosure percentage and any language around the SuperVision attach rate — that is the product family designed to grow out of single-OEM dependence.

Long-term (1–3 years): The structural question is whether NVIDIA DRIVE Hyperion stays inside Hyundai vehicles after 2029 or whether the in-house stack displaces NVIDIA too. Park Min-woo's Sep 12 framing — using NVIDIA data to \"train and refine\" Hyundai's own platform — implies NVIDIA is the on-ramp, not the destination. If Hyundai reaches 5.55M vehicles by 2030 with even half on its proprietary Level 2 stack, the ADAS silicon market share shifts toward whoever keeps the lowest-cost-per-TOPS bid for the post-2029 refresh — currently Qualcomm Snapdragon Ride Flex and the next-generation NVIDIA Thor.

The cleanest trade is not on Mobileye. It is on Hyundai Mobis: one KOSPI-listed entity collects the integration margin on every chip family that wins a Hyundai ADAS slot — EyeQ today, DRIVE Hyperion in 2028, Snapdragon Ride Flex in parallel, and whatever silicon rides inside the proprietary 2029 stack.

Investable takeaway

MMobileye GlobalMBLY--
--Vol --
-
Bearish
  • Hyundai Group's late-2029 in-house Level 2 timeline plus 2028 NVIDIA bridge structurally erodes the 69% one-customer concentration disclosed in the Q2 2026 10-Q.
  • $24.5B 8-year pipeline is the offset, but Q2 2026 revenue was already flat YoY at $508M with FY2026 guide 4–7% — the cliff starts in 2027, the offset arrives in 2028+.
  • At 2.7x EV/Sales TTM the multiple already prices distress; further compression requires visible quarterly share loss to NVIDIA or Qualcomm rather than just the Sep 12 narrative.
NNVIDIANVDA--
--Vol --
-
Bullish
  • 2028 DRIVE Hyperion Level 2 deployment at Hyundai Motor Group (one of the world's top-three OEMs by volume) adds incremental revenue on top of $2.35B FY2026 Automotive segment.
  • Auto is still <1% of NVIDIA's $96.2B Q2 FY2027 total — absolute dollars meaningful, but the multiple is already 17.5x EV/Sales TTM, so this is a confirmation, not a re-rating catalyst.
  • Hyundai's stated intent to use NVIDIA data to train its own platform keeps DRIVE Hyperion in the 2028–2029 cohort even as the proprietary stack matures — durability of the slot is the watch item.
QQualcommQCOM--
--Vol --
-
Mixed
  • Snapdragon Ride Flex + Hyundai Mobis agreement (Jan 7, 2026) is the strongest claim to the silicon inside Hyundai's 2029 in-house stack — auto revenue >$5B annualized in Q2 FY2026 supports the R&D.
  • $15B ADAS pipeline and $4B 2026 car-revenue target are credible, but losing the 2028 Level 2 Hyundai platform to NVIDIA DRIVE Hyperion caps the near-term attach rate.
  • Trade at 4.6x EV/Sales TTM — auto growth is a multi-year compounder, not a single-quarter catalyst.
0Hyundai Mobis012330.KS--
--Vol --
-
Bullish
  • Captive Tier-1 integrator for Hyundai Motor Group; collects bill-of-materials margin on EyeQ today, DRIVE Hyperion in 2028, and Snapdragon Ride Flex in parallel — the chip winner matters less than the integrator.
  • Q2 2026 revenue KRW 16.3T (+2.4% YoY) per Aug 14, 2026 earnings summary; ADAS content per vehicle rises regardless of who supplies the silicon.
  • 5.55M Hyundai Group 2030 sales target (per August 2026 Investor Day) means a wider TAM for in-house software + Mobis integration through the 2028–2030 window.
0Hyundai Motor005380.KS--
--Vol --
-
Mixed
  • Long-term: owning the ADAS stack preserves software-margin dollars and OEM bargaining power vs. Mobileye, NVIDIA, and Qualcomm — a strategic positive for the 5.55M 2030 unit target.
  • Short-term: a two-year in-house delay (now 2029 vs. originally 2027) raises R&D burn and widens the gap to Tesla and Chinese OEMs, per the Reuters Sep 12 read.
  • Jan 7, 2026 +14.9% record-high print priced in the NVIDIA partnership narrative; the Sep 12 in-house disclosure is the next data point the buy-side has to absorb.
0Kia Corporation000270.KS--
--Vol --
-
Watch
  • Joint signatory of the March 16, 2026 NVIDIA partnership with Hyundai Motor — captures the same 2028 DRIVE Hyperion Level 2 pipeline and the 2029 in-house stack transition.
  • No Mobileye-specific disclosure on Kia's share of the 69% Mobileye major-customer concentration, but Kia vehicles share the Hyundai Group ADAS architecture.
  • Q2 2026 Hyundai Group global retail of 999K vehicles includes Kia — direct read-through to Kia's per-unit ADAS content.
TTaiwan SemiconductorTSM--
--Vol --
-
Watch
  • Fabs incumbent EyeQ chips and is the leading foundry candidate for both NVIDIA's next-gen DRIVE Thor and Qualcomm Snapdragon Ride Flex successors — net auto-foundry volume is directionally neutral.
  • Mobileye chip attrition (Samsung also a Mobileye supplier) is small relative to total TSMC revenue; ADAS is a tiny share of foundry mix.
  • Watch for changes in Mobileye's disclosed wafer partner on future 10-Q filings as a leading indicator of mix shift.
IIntelINTC--
--Vol --
-
Watch
  • Still controls ~96.9% of Mobileye voting power (50M Class A + all Class B shares per the 2026 proxy); Mobileye's strategic setbacks are Intel's indirect exposure.
  • Mobileye was a flagship IDM 2.0 automotive proof point; the Hyundai attrition weakens that narrative without affecting Intel's core foundry or CPU lines materially.
  • Mobileye market cap of ~$6.7B is a small fraction of Intel's enterprise value — direct P&L impact is limited; the reputational read-through is the watch item.