Micron completed long-term memory and storage agreements with Qualcomm and six automotive technology suppliers. The contracts extend a much larger take-or-pay program, but Micron has not disclosed the seven deals’ volumes, price terms, values or capacity allocations.
Micron’s latest automotive agreements extend a consequential shift toward long-term, take-or-pay memory contracts. They do not amount to a newly disclosed $100 billion automotive order: Micron has not said how much product the seven customers committed to buy, what they will pay or which factories will supply them.
Micron said in its July 16 announcement that it had completed Strategic Customer Agreements with Qualcomm, Visteon, HARMAN, JOYNEXT, DENSO, Astemo and Hyundai Mobis. The group consists of a chip-platform provider and automotive technology suppliers rather than vehicle manufacturers.
That distinction broadens Micron’s customer map. The announcement connects the counterparties to cockpit, driver-assistance, connectivity and other vehicle systems. HARMAN is also a Samsung Electronics unit, according to a market report.
Micron says the agreements give both sides more visibility into production planning, pricing and future product requirements. That visibility can support the lengthy qualification work and supply continuity demanded by vehicle programs with long production lives.
The announcement stops before the commercial terms. It provides no contract-specific duration, product list, volume, price formula, minimum value, customer deposit or factory allocation. It also does not say whether the agreements are exclusive.
Micron supplied the important contract mechanics in its fiscal third-quarter prepared remarks. The company said it had completed 16 agreements across data-center, consumer and automotive markets. It described all 16 as take-or-pay contracts with binding commitments to purchase specified volumes.
The typical agreement runs from 2026 through 2030, but automotive contracts generally run for three years. Micron classified four customers as very large and three as medium-sized; it described the remaining automotive customers as smaller but did not map the seven July names to those size categories. Across the full program, the signed agreements represented roughly 20% of Micron’s DRAM volume and one-third of its NAND volume over the relevant periods.
The price protection also needs a narrower reading than the July 16 announcement suggests. Micron said its largest agreements generally have both a ceiling at the calendar second-quarter market price for existing products and a floor through the contract term. Several agreements accounting for a modest share of contract revenue instead have fixed prices or no price band, leaving prices subject to the market. Micron did not identify the structure used for any of the seven automotive counterparties.
For agreements signed by the time of the June remarks, including contracts executed after the quarter ended, Micron estimated approximately $100 billion of remaining performance obligations. That estimate was based on minimum committed volumes and minimum prices; the company said it was not a forecast of total revenue. The remarks do not allocate any of it to the seven companies named in July.
Micron also projected $22 billion in customer deposits and related financial commitments across the signed portfolio, including about $18 billion of cash deposits. Those deposits are not free cash flow or permanent financing: Micron said they would appear as financing cash flows and be returned to customers toward the latter half of the agreements. They put customer capital behind future purchases, but they are neither automotive revenue nor a valuation for the July deals.
The automotive business is growing, though it remains much smaller than Micron’s data-center operation. In the same quarter, Micron reported $4.6 billion of Automotive and Embedded Business Unit revenue, or 11% of company revenue, compared with more than $25 billion from data centers. Automotive and embedded revenue rose 71% sequentially, driven by higher prices and bit shipments, and the unit’s gross margin reached 79%.
Micron estimates that vehicles with Level 2+ driver assistance or greater autonomy carry more than five times the memory and storage content of an average vehicle. It expects those vehicles to exceed 20% of the mix in 2026 and 40% by 2030. Those are company forecasts, not volumes guaranteed by the new agreements.
The immediate leverage comes from scarce supply. A contemporaneous report said strong memory demand had allowed Micron, SK Hynix and Samsung Electronics to command premium prices. It also identified Micron as the only U.S.-based producer of high-bandwidth memory used with Nvidia AI processors. The seven automotive agreements, however, do not name HBM or any other product, so data-center HBM leadership should not be treated as their product scope.
Micron expects both DRAM and NAND supply-demand conditions to remain tight beyond 2027. It says new supply is constrained by long fab construction schedules, shortages of skilled trades, permitting, energy infrastructure and the rising manufacturing burden of HBM, which also pressures non-HBM output. Long-term contracts can decide who receives scarce supply and give Micron confidence to invest; they do not create that supply on signing.
Two agreements announced earlier in July show what Micron discloses when it is willing to be more specific. On July 1, Micron and General Motors announced an agreement covering LPDRAM, NOR and UFS NAND. Micron linked that supply to a $2 billion modernization of its Manassas, Virginia, fab, which the company said had begun production earlier in 2026.
On July 6, Micron and Ford announced a separate long-term agreement for next-generation vehicle production. Micron said it was increasing output of key automotive memory products and cited advanced DRAM expansion at Manassas, but it did not disclose Ford’s volumes or contract value.
Both agreements are among the same group of 16 discussed in June. They demonstrate that Micron’s automotive contracting reaches automakers as well as their technology suppliers, while also exposing the information missing from the July 16 release: named products and at least some connection to physical capacity. The evidence does not establish that the seven later agreements use the same products or depend on Manassas.
An item published at 11:15 a.m. Eastern on July 16 said Micron shares were down 5.13% at $857.95 on Nasdaq. That was an intraday observation, not a closing price, and neither that report nor the other retained sources connected the decline to the automotive agreements.
The market was already divided over whether high memory profits can endure. A pre-announcement analysis described Micron and Sandisk as the two biggest blue-chip gainers in the first half but said their earnings multiples remained in the S&P 500’s lowest 20%. It attributed the discount to memory’s history of booms and busts.
That analysis also presented the countercase the provisional framing missed: Bernstein analyst Mark Newman argued that the multiples priced in an imminent profit collapse despite continuing AI data-center demand, while D.A. Davidson analyst Gil Luria said investors had not grasped how AI had changed the memory business. The automotive contracts offer a possible stabilizer through committed volumes, but their undisclosed terms provide no basis for deciding which view they strengthen.
The central question is no longer whether automotive customers want supply assurance. It is whether these automotive contracts materially change Micron’s revenue durability, capital planning or exposure to a future memory downturn.
Answering it requires contract-specific or at least segment-level evidence: committed automotive volumes, the applicable price structure, deposits and remaining performance obligations attributable to automotive customers, product qualifications, capacity allocations and the vehicle programs that reach production. Actual purchases and margins over the agreements’ terms will then show whether they created incremental demand or mainly reserved supply those customers would have sought anyway.
OpenAI has put conversations and Projects back in its redesigned ChatGPT desktop app and enabled cloud Work threads to move across devices, correcting the launch's biggest usability failures without merging local Work or Codex histories.
Meta reportedly plans to put departing AWS compute executive Dave Brown to work on its data-center buildout, adding hyperscale operating experience while leaving any customer-facing cloud business conditional and undefined.
China has paired a five-year AI training offer for developing countries with cooperation centers, a weather-warning rollout and a new 29-country organization. The package gives Beijing a platform for influence, but no budget, selection rules or delivery timetable has been published.
RoboTTT extends a robot policy’s visuomotor context to 8,000 timesteps and raises its average rubric score from 42% to 79%, but the evidence comes from three author-run assembly tasks and the longest task was completed in only two of ten trials.
Zhipu reportedly reached $1 billion in annual recurring revenue in July, roughly four times a March estimate, but the unconfirmed run rate is not annual sales and still sits far ahead of recognized cloud revenue while margins remain thin.
Nebius has arranged its first senior secured facility against an operating GPU deployment and one customer’s cash flows. The deal adds project-level debt to its expansion toolkit, but an unnamed customer and a larger rival financing limit what it proves about the rest of Nebius’s backlog.
Cursor says automatically running a repository-root git.exe on Windows does not meet its criteria for patching, while the researcher calls it an untrusted-search-path defect and separate research shows the same weakness across several competing AI coding tools.
Huawei publicly displayed a 16-cabinet Atlas 950 configuration rated at 1 EFLOPS in FP8, providing tangible evidence of its system-scale AI strategy while leaving price, power use and sustained workload performance undisclosed ahead of the full system's planned fourth-quarter release.
CIA Director John Ratcliffe said US intelligence is consistent with an estimate that Russian recruits last 20 to 30 minutes on Ukraine’s battlefield, but the public trail leads to an unsourced claim about assault troops and does not establish a representative average.
UK testing places leading open-weight models four to seven months behind selected closed-model cyber results, yet longer attack chains, U.S. benchmarks and mixed cost comparisons show why that interval is a warning signal rather than a universal capability clock.
TSMC reached the top of its second-quarter revenue guidance and raised its 2026 outlook, but a one-off investment gain boosted profit growth while the company committed more capital to 2-nanometer production, advanced packaging and an undated Arizona expansion.
Moonshot AI's Kimi K3 added to a global technology selloff with near-frontier performance, but unreleased weights, mixed cost comparisons and a recommended 64-accelerator deployment leave its effect on chip demand unresolved.
Moonshot AI has made Kimi K3 available through its apps and API, pairing a 2.8-trillion-parameter architecture with early frontier-level results, but the model's open-weight claim cannot be tested until its weights and technical report arrive.
Two binding EU decisions require Google to give rival AI services comparable access to 11 Android features and offer eligible search competitors a restricted, anonymized dataset, but phased deadlines, certification, pricing and privacy safeguards leave the competitive effect unproven.
Databricks has signed a term sheet for a Coatue-led financing at a $188 billion valuation, while unidentified sources put the round at $3 billion. The proposed capital would deepen its push into AI governance, data agents and operational databases, but the transaction remains open and the company supplied no new operating figures.
Netflix says generative AI workflows were used on roughly 300 titles in 2026, but its only quantified example tied to that disclosure covers 17 minutes and does not establish how much finished material, spending or labor changed across the slate.
Gemini 3.5 Pro missed its expected June rollout. An anonymous-source account says a late-June data change fell short of Google's coding goals, but Google has confirmed only partner testing—not the reported cause, a new date, or public results and pricing.
Google is bringing Gemini Omni editing and a reusable face-and-voice avatar to Vids, but the sharper distinction is account-level identity across Vids and Gemini rather than a new category of AI video; Vids already offered Veo generation and customizable avatars, while specialist rivals already sell digital presenters.
Intel plans to adopt Gemini Enterprise across engineering, supply-chain and corporate work while extending chip-simulation capacity into Google Cloud, but the companies supplied no rollout schedule, contract value or performance baseline.
The Navy has approved an immediate department-wide framework for turning data into operational effects, but the public rollout leaves budgets, owners, deadlines, performance measures and assurance rules to a promised implementation roadmap.