brain/
← all entities
entityAAPLstock-market

Apple

Notes

Apple

One-line summary: TSMC's largest leading-edge customer for over a decade; reached a preliminary deal with Intel for some chip production (WSJ, May 2026) — first publicly reported non-TSMC supplier for Apple silicon in years, with the US government as active broker.

Valuation (September 10, 2026)

Last: $315.34 at the Wednesday 2026-09-09 regular close (−0.28%). From Twelve Data (twelvedata). 52-week $225.95–$344.57 (−8.5% off high). Surprise and shine ran Sep 9 (iPhone Duo hardware; not a disclosed aggregator toll). Prior vault last on this page was the Sep 2 close $324.96.

Launch-day tape (2026-08-25, not the vault last): AAPL $309.90 −0.14% vs QQQ $710.72 +0.62% (stockanalysis AAPL; stockanalysis QQQ; Yahoo matches the Aug 25 AAPL close 309.90). One-day observation, not a sizing input.

Gage's judgment (as of 2026-08-31)

Gage, 31 Aug 2026. Extra upside vs the iPhone/services multiple is about a few percent over 6–12 months, and only if a later print credits desktops and high-memory mix (not laptops). If that never shows, we lose nothing extra from this story. Gage's judgment, not grain / not issuer fact. Last-price as-of stays the vault last: $316.85 (Monday 2026-08-31 regular close). Friday $319.70 is prior. Launch day 2026-08-25 was not a spike (AAPL lagged QQQ). Street AI frame stays iPhone/Siri/services. The 10 paper shares are Paul's directed ticket, not this thesis — do not treat the Mac-local-AI story as the reason for those shares. Watch: Sep 22 mini ship, late-Oct 512GB Studio, next 10-Q MD&A desktop vs laptop. Sep 9 iPhone event is the actual stock event this month. Not a buy/sell/size. Not a paper sleeve. No source URL — do not invent one.

What it is

SoC designer (M-series, A-series) for iPhones, iPads, Macs, and Apple silicon for adjacent products. Has been a TSMC-exclusive customer at leading nodes for the M and A series. Ships ~200M iPhones annually plus tens of millions of iPads and Macs.

Why it matters to stock-market

If Apple confirms even a fraction of its silicon shifting to Intel Foundry, that is the highest-conviction signal possible for the us-fab-capacity-bottleneck thesis — Apple has been the foundry industry's most discriminating customer, and a defection sets a precedent for other hyperscalers. The deal also has unusual political durability (Trump-administration brokered).

Key facts

  • Apple-Intel deal status (May 2026): WSJ reported "preliminary agreement" (May 5–8, 2026); most credible framing is a formal agreement-in-principle, products/volume/timeline TBD. Neither Apple nor Intel has made a public statement. TheStreet's "signed" / GigaNectar's "sealed" language overstates the sourcing.
  • Node and volume specifics (Kuo/Pu, WCCFTech, NotebookCheck): M7 on Intel 18A-P — ~15-20M units/yr, production start 2027, targeting MacBook Air + iPad Pro. A21 iPhone chip on Intel 14A — ~60M units/yr (~20% of total iPhone production), mass production 2028. The A21 is ~80% of the deal's total order value — it is the load-bearing revenue event for Intel IFS. Apple's lifecycle framing: "small-scale testing 2026, ramp 2027, continued growth 2028, decline 2029." Apple is positioning Intel as a secondary foundry for supply chain diversification, not TSMC replacement. From 2026-05-21-autoresearch-apple-intel-deal-may-21-2026.
  • TSMC response (Bernstein): Apple's Intel volumes are "too small to matter" for TSMC. Apple represents ~25% of TSMC total revenue; 20% iPhone shift to Intel 14A reduces TSMC's Apple share proportionally, but TSMC fills from AI/HPC customers immediately. From 2026-05-21-autoresearch-apple-intel-deal-may-21-2026.
  • Apple's stated rationale (Q2 FY2026 earnings, April 2026): Tim Cook stated primary constraint is "the availability of the advanced nodes their SoCs are produced on, not memory" — driven by AI chip demand and local-AI device requirements. Cook did NOT name Intel on the call; supply constraint is indirect corroboration.
  • Intel Q1 call (April 23, 2026): Google was named as a new anchor customer; Apple was NOT named. CEO Tan said "if Intel does not secure an anchor customer by late 2026, the likely outcome is a slower path forward" — making Apple the implied make-or-break anchor.
  • Government role: Trump White House actively brokered the deal; Trump personally advocated for Intel in White House meeting with Cook.
  • Samsung parallel evaluation: Apple executives physically visited Samsung's Taylor, Texas facility (Bloomberg, May 5, 2026). Apple has a preliminary Intel agreement but Samsung is an active second evaluation — not a distant fallback. See samsung.
  • Market reaction: INTC up ~14% on Apple talks reporting (May 5 2026); INTC up ~240% YTD as of May 2026.

Strengths (from a thesis-input perspective for INTC long)

  • Apple's TSMC-only stance breaking is hard to reverse politically
  • Cook's public capacity-constraint statement gives Apple cover to diversify
  • White House active brokering reduces "deal falls apart in 6 months" risk

Weaknesses (from a thesis-input perspective)

  • Deal is preliminary, no orders yet — reporting could prove premature
  • Initial scope is low-end M-class, not iPhone — incremental, not transformative
  • Apple is also evaluating Samsung — Intel may share or lose the spoils

OpenAI ChatGPT integration dispute (Mar/May 2026)

  • nick-turley (Head of ChatGPT) in 2026-03-15-bg2-chatgpt-super-assistant-era indirectly references the Apple partnership context: "There's a couple of big partnerships last year, Apple, Reliance with Gemini. Those are two big user bases." As of March 2026, the Apple integration was still characterized as a partnership in good standing.
  • By May 2026 (per 2026-05-15-all-in-podcast-trump-xi-benioff-saaspocalypse-openai-apple, previously cited on openai's page): OpenAI is considering suing Apple over the Siri / iOS ChatGPT integration "having gone so poorly." Apple's defense includes annoyance that OpenAI is building hardware to compete (Jony Ive). david-friedberg's framing: Apple may need to either build its own AI or white-label with anthropic.
  • AAPL implications:
    • Distribution downside. Apple's lack of a credible first-party AI is a competitive vulnerability — Google has Gemini, Microsoft has OpenAI partnership, Meta has Llama. If the OpenAI integration ruptures, Apple either delays or pivots to a different partner (Anthropic — see anthropic page for the Friedberg framing). Either path delays Apple's AI roadmap by ~12-24 months.
    • Hardware-competition signal. Jony Ive working with OpenAI on consumer hardware (a non-phone form factor) is a strategic threat to iPhone product cycle defensibility. Apple's response so far is friction-based (slowing the integration, withholding API access).
    • Cross-link to ChatGPT-as-discovery-channel. nick-turley in the same source discusses ChatGPT-shopping as a use case OpenAI is investing in. If ChatGPT becomes the dominant product-discovery surface for many users, Apple loses both the AI layer of the stack AND part of the App Store / commerce flywheel attached to it. The ChatGPT-shopping development is a slower-burn AAPL concern than the integration dispute.

Apple sues OpenAI + the on-device local-model "screaming buy" thesis (2026-07-18)

Two developments from 2026-07-18-podcast-all-in-podcast-can-the-ai-industry-regulate-itself-stripe-wants (multi-context):

  • The OpenAI dispute went to litigation. On July 10, 2026 Apple filed a 41-page lawsuit against openai alleging stolen trade secrets — that OpenAI used Apple IP for its consumer hardware device, poached ~400 Apple employees, and that its chief hardware officer (ex-Apple iPhone-design VP "Tang 10") solicited "actual parts" from Apple interviewees. Resolves the May "who sues whom" ambiguity: Apple moved first. Bearish for the ChatGPT-default-Siri distribution path; a spur to Apple first-party/on-device AI.
  • On-device local models as the AAPL bull case (jason-calacanis, speaker-attributed): with Mark Gurman reporting an M7 Ultra supporting ~1.5TB RAM (double current), "an opus level model running on your Mac studio," Calacanis calls Apple "a screaming buy": "they will put so much downward pressure on Claude and OpenAI by just putting local models and supporting them with this memory architecture... employers go, oh, I can just run 90% of my workloads on the local Mac studio." This is the AAPL expression of edge-inference-shift — unified-memory Apple silicon as the enterprise/consumer local-inference substrate, a slow-burn demand vector distinct from (and partly opposed to) the cloud-AI-laggard framing. ⚠ Interested/promotional ("not financial advice") and unverified on the M7 Ultra spec.

AI aggregator-toll thesis (2026-07-14)

A distinct, non-consensus AAPL thesis surfaced on 2026-07-07-podcast-the-compound-and-friends-leveraged-etf-casino-apple-breaks-out-sentiment: rather than an AI laggard, Apple as the AI aggregator/tollbooth — its ~2.5B-device base + an iOS27 default-model selector let it monetize everyone else's frontier model (josh-brown: "Apple is the tollbooth… they're going to get paid coming and going regardless of which AI the consumer wants to use. That is the checkmate"; Dan Ives $400 target, ~$15B/yr AI-services opportunity). Near-term earnings driver already working: iPhone 17 momentum + memory-cost passthrough (DRAM/NAND +~98%/quarter; iPhone 18 Pro possibly +$200) "taken to margin" (lumberflation analogy). Filed as a hypothesis with the load-bearing toll-payment mechanics flagged unverified: apple-ai-aggregator-toll-to-services-rerate.

M6 Mac mini / M5 Ultra Studio launch (2026-08-25)

Attach-only. Local-AI Mac boxes ≠ an AAPL AI re-rate. Not a trade. No buy/sell/size. From 2026-08-31-m6-mac-mini-m5-ultra-studio-is-a-local-ai-hardware-datapoint.

  • Launch (confirmed, Newsroom, 2026-08-25): M6 in the new Mac mini and M5 Ultra in the new Mac Studio. Same-day product releases: Mac mini with M6 and M5 Pro (pre-order Aug 25; availability beginning September 22); Mac Studio with M5 Max and M5 Ultra (pre-order Aug 25; arriving September 22). Quote: "Mac Studio with 512GB of unified memory is coming in late October." Do not flatten "Studio ships Sep 22" into "512GB ships Sep 22."
  • SKUs (Newsroom marketing, not third-party benches): M6 mini starts at $899 — 12-core CPU, 12-core GPU, Dual 16-core Neural Engine, 16GB standard unified memory configurable up to 32GB, up to 170GB/s. M5 Pro mini starts at $1,699 — up to 64GB, 307GB/s. M5 Max Studio starts at $2,499 (up to 128GB). M5 Ultra Studio starts at $5,499 — up-to-36-core CPU, up-to-80-core GPU, up to 512GB unified memory, 1.2TB/s. Chip-page claims: M6 is Apple's first 2 nm chip; M5 Ultra is the first M-series quad-die UltraFusion SoC and "lets users … run huge LLMs with hundreds of billions of parameters entirely on device."
  • FY26 Q3 Mac mix (10-Q dollars, not a printed percentage): Form 10-Q period ended June 27, 2026 Note 2: Mac $10,352 million of $109,417 million total. Same dollars in the July 30, 2026 8-K Exhibit 99.1 and the Newsroom condensed statements PDF. Stamp $10,352m / $109,417m. Arithmetic 10,352 / 109,417 = 9.46% (derived). Do not stamp "9.5%" as a 10-Q figure. Mac grew 29% YoY ($8,046m in Q3 FY25) and is still a high-single-digit slice.
  • MD&A (same 10-Q, Mac paragraph): "Mac net sales increased during the third quarter and first nine months of 2026 compared to the same periods in 2025 due to higher net sales of laptops." No desktop / mini / Studio attribution. The Q3 print predates the 2026-08-25 desktop launch (period ended June 27; 8-K July 30). Do not attach this launch to Q3 Mac growth.
  • Local-AI Mac boxes ≠ AAPL AI re-rate. Cook's Newsroom / 8-K Ex. 99.1 lead is WWDC26 "all-new Siri AI" plus double-digit iPhone, Mac and Services. Mac is named as a hardware revenue record, not as the AI-equity story. Desktop local-AI SKUs do not re-rate AAPL and do not retarget apple-ai-aggregator-toll-to-services-rerate onto Mac boxes. The Calacanis M7 Ultra ~1.5TB "screaming buy" block above stays podcast judgment — this launch is M6 / M5 Ultra / 512GB, not M7 Ultra / 1.5TB.

Surprise and shine — Sep 9, 2026 pre-event brief

Pre-event brief: confirmed invite vs rumor SKUs. Do not flatten the M6 Mac mini / Studio clip into this event. Not a trade. No buy/sell/size. From 2026-09-01-aapl-pre-event-brief-surprise-and-shine.

  • Event (confirmed, Apple paper): Developer News, Aug 26; Apple Events. Tagline Surprise and shine. Wednesday, September 9, 2026, 10:00 a.m. PT. Livestream: apple.com and Apple TV. Developer News also names YouTube Live; the Events page does not. Invite / announcement date August 26, 2026. Apple named no SKUs.
  • Issuer-unconfirmed (do not stamp as of the pre-event brief): iPhone 18, foldable, Watch, Siri as Sep 9 products. Apple's pages called it “a special Apple Event” and stopped there. Closed 2026-09-09 — see launch section below.
  • Ternus CEO: April 20 Newsroom — John Ternus CEO effective September 1, 2026. The invite does not name him. “First under CEO Ternus” is chronology from two confirmed dates, not an Apple event-page sentence.
  • Venue: Steve Jobs Theater / Apple Park is journalist-listed (The Verge; PCMag; MacRumors), not on fetched Apple pages. Do not stamp the theater as issuer-confirmed.
  • Siri AI is Apple paper for WWDC, not for Sep 9 (Events page WWDC June 8, 2026; 10-Q MD&A; Cook Q3 Newsroom / 8-K Ex. 99.1).
  • Memory / pricing (issuer primaries, not iPhone 18 +$200): Cook on the July 30, 2026 call confirmed iPad/Mac price rises and “For September, we expect to pay even higher memory costs.” He did not print an iPhone 18 Pro dollar hike. DRAM/NAND ~98% and iPhone 18 Pro possibly +$200 stay prior podcast/hypothesis on this page’s aggregator-toll block — not this clip’s grain.
  • Rumor layer (named journalist, not grain): Bloomberg via Yahoo (foldable / 18 Pro/Max / spring delay); MacRumors expected SKUs; Verge listed theater + rumor Pro/foldable; PCMag WSJ $1,299 estimate — WSJ not fetched, not this grain.
  • Gage: this clip is empty. Existing Gage (Mac-local-AI few percent; 10 paper shares are Paul’s directed ticket, not this thesis) is unchanged. Do not invent a new range.

Surprise and shine — Sep 9, 2026 launch (hardware, not a toll)

Event ran. Hardware SKUs printed. Conversion bar for apple-ai-aggregator-toll-to-services-rerate is a disclosed frontier-lab aggregator payment — not met. Stay hypothesis. Do not graduate. Not a trade. From 2026-09-10-research-event-day-gates.

  • iPhone Duo (Newsroom): foldable, starts at $1,999. Inner 7.6" / outer 5.4". A20 Pro. Pre-order October 16; ships October 23.
  • Also launched: iPhone 18 Pro / Pro Max; Watch Series 12 / Ultra 4; AirPods 5.
  • Siri AI beta with iOS 27 on September 14. Copy is on-device Apple Intelligence + Private Cloud Compute. Footnote 2: expanded access to server-side features "will be available for a fee in the future" — an Apple usage meter, not a disclosed default-placement payment from a frontier lab.
  • Hardware + a future Apple fee ≠ the aggregator-toll conversion. Do not re-date AAPL.

Open questions

  • apple-intel-deal-finalization — will it convert, on what products, on what timeline?
  • apple-ai-aggregator-toll-to-services-rerate — does the install-base aggregator toll re-rate AAPL on AI-services value it captures without building a frontier model?
  • OpenAI partnership resolution. Does Apple settle the integration dispute, switch to Anthropic, or accelerate first-party AI? Resolution path determines whether the Apple ecosystem becomes an AI-integrated stack or remains an AI-pass-through. Cross-link openai and anthropic for the two competing-partner thesis paths.

Sources

Related

Referenced by
Mechanisms
AI memory crowd-out → non-LTA buyers are the residual claimants → device units fall but revenue holds → the consumer dollar transfers from OEM to memory supplierHBM pre-committed + CoWoS half-owned by Nvidia → DRAM spot price +158% → memory/packaging bottleneck downstream of foundrySamsung SF2P yields hit 70% + Taylor 90% ready + Apple visit → genuine third foundry leg → order flow splits, not concentratesTSMC saturation → 4-year price hikes → marginal buyers shift → Intel anchor stack → re-rate
Concepts
AI safety as a vector for regulatory capture / centralizationFrontier-class intelligence is escaping the datacenter to the edgeUS fab capacity bottleneck → Intel as the alternative foundry
Questions
Does Apple's 2.5B-device install base + an iOS default-model selector let it toll the frontier-LLM winners regardless of which one wins — re-rating AAPL on AI-services value it captures without building a frontier model?Does Samsung have a credible leading-edge foundry comeback that would weaken the Intel-as-alternative thesis?F1 US viewership growth → Apple US TV rights deal → FWONK broadcast revenue re-rateWill the Apple-Intel deal convert to actual orders, on which products, on what timeline?
Entities
Ferrari (RACE)IntelLiberty Media (Formula One Group — FWONK / FWONA)Michael BatnickOpenAISamsung FoundryTSMC (Taiwan Semiconductor Manufacturing Co.)