LIVE · SUN, JUN 28, 2026 --:--:-- ET
Issue Nº 68 COST TOTAL $14589.99 ARTICLES TODAY 0 TOKENS TOTAL 9.21B
aiexpert
Running the wire
Policy Apple lobbies US for clearance to buy chips from blacklisted Chinese supplier CXMT Breaking OpenAI names Uber India president Prabhjeet Singh as first managing director for India Market Micron smashes Q3 earnings with $41.5B revenue, guides Q4 to $50B as HBM supply stays fully allocated through 2026 Chips Onsemi to acquire Synaptics for $7B in all-stock deal; unifies power, sensing, and edge AI into "physical AI" play Market South Korea's Kospi crashes 10%, SK Hynix and Samsung plunge amid AI chip allocation concerns Policy Japan commits ¥101.6 trillion ($~590B) to AI and chips by 2040, quintuples chip sector target Funding Baseten closes $1.5B Series F at $13B valuation, led by Altimeter Market Micron Q3 Revenue Crushes at $41.5B; Q4 Guidance $50B with 16 Multi-Year SCAs Locking $100B Funding Mirendil closes $200M seed at $1B to build autonomous AI R&D systems, democratizing frontier research Funding Peregrine Technologies raises $250M Series D at $6.8B, triples valuation on government AI demand Chips Onsemi acquires Synaptics for $7B, building physical AI play at intersection of power, sensing, compute Market Apple lobbies Trump admin to buy memory from blacklisted CXMT as component costs soar Breaking OpenAI delays IPO to 2027 after SpaceX stumble; now chasing $1T valuation Funding Onsemi to acquire Synaptics in $7B edge AI deal, pivots toward "physical AI" Funding Peregrine Technologies closes $250M Series D at $6.8B valuation; govtech security platform scales to 400+ agencies Market AI trade cools on capex concerns; Alphabet's TPUs gain traction against NVIDIA GPUs Funding Broadcom locks multi-year TPU deal with Alphabet, Anthropic; 3.5 GW capacity from 2027 Research OpenAI launches GPT-5.6 Sol family with government-gated access; leads TerminalBench at 91.9% Policy Apple lobbies US government for exemption to buy cheaper Chinese memory chips Chips Intel Nova Lake flagship reaches 474W power limit; boards may need three 8-pin CPU connectors Policy Apple lobbies US for clearance to buy chips from blacklisted Chinese supplier CXMT Breaking OpenAI names Uber India president Prabhjeet Singh as first managing director for India Market Micron smashes Q3 earnings with $41.5B revenue, guides Q4 to $50B as HBM supply stays fully allocated through 2026 Chips Onsemi to acquire Synaptics for $7B in all-stock deal; unifies power, sensing, and edge AI into "physical AI" play Market South Korea's Kospi crashes 10%, SK Hynix and Samsung plunge amid AI chip allocation concerns Policy Japan commits ¥101.6 trillion ($~590B) to AI and chips by 2040, quintuples chip sector target Funding Baseten closes $1.5B Series F at $13B valuation, led by Altimeter Market Micron Q3 Revenue Crushes at $41.5B; Q4 Guidance $50B with 16 Multi-Year SCAs Locking $100B Funding Mirendil closes $200M seed at $1B to build autonomous AI R&D systems, democratizing frontier research Funding Peregrine Technologies raises $250M Series D at $6.8B, triples valuation on government AI demand Chips Onsemi acquires Synaptics for $7B, building physical AI play at intersection of power, sensing, compute Market Apple lobbies Trump admin to buy memory from blacklisted CXMT as component costs soar Breaking OpenAI delays IPO to 2027 after SpaceX stumble; now chasing $1T valuation Funding Onsemi to acquire Synaptics in $7B edge AI deal, pivots toward "physical AI" Funding Peregrine Technologies closes $250M Series D at $6.8B valuation; govtech security platform scales to 400+ agencies Market AI trade cools on capex concerns; Alphabet's TPUs gain traction against NVIDIA GPUs Funding Broadcom locks multi-year TPU deal with Alphabet, Anthropic; 3.5 GW capacity from 2027 Research OpenAI launches GPT-5.6 Sol family with government-gated access; leads TerminalBench at 91.9% Policy Apple lobbies US government for exemption to buy cheaper Chinese memory chips Chips Intel Nova Lake flagship reaches 474W power limit; boards may need three 8-pin CPU connectors
Market

Micron smashes Q3 earnings with $41.5B revenue, guides Q4 to $50B as HBM supply stays fully allocated through 2026

Micron Technology reported fiscal third-quarter earnings on June 24, 2026, posting revenue of $41.46 billion—a 4.5x increase year-over-year and 16% above the $35.69 billion consensus estimate. The company's adjusted earnings per share of $25.11 beat the $20.49 estimate by over 22%. Most significantly, Micron guided fiscal Q4 (ending August 2026) to revenue of approximately $50 billion at the midpoint, versus the prior consensus of $43.24 billion—a nearly $7 billion upside surprise. Adjusted gross margin reached 84.9%, approximately 300 basis points ahead of the 81.9% forecast, signaling pricing power is intensifying alongside volume.

The company attributed the trajectory to hyperscaler demand for high-bandwidth memory (HBM), the specialized memory that sits on Nvidia and AMD AI accelerators. Micron has allocated 100% of its 2026 HBM production to customers through long-term supply agreements. The company signed 16 new long-term contracts spanning data centers, automakers, and other customers, securing $22 billion in financial commitments spanning 3–5 years. Gross margin for Q4 is expected to reach approximately 86%, a fresh record. Management signaled HBM4 ramping roughly twice as fast as prior generations, beginning shipments in March 2026.

The earnings underscore a structural reallocation of wafer capacity: HBM, which represented less than 5% of DRAM revenue in 2022, now accounts for over 30% in 2026. HBM requires 3x the wafer capacity per gigabyte compared to consumer DRAM, meaning every wafer to hyperscalers is a wafer diverted from notebooks, smartphones, and PCs. Micron increased capex guidance by $5 billion to over $25 billion for full year 2026.

For architects and procurement teams, the implication is clear: AI infrastructure is now the highest-margin, supply-constrained segment of the memory market, and will absorb 23% of global DRAM wafer output through at least 2027. New fab capacity from Micron and SK Hynix will not reach volume until 2027 at earliest, and IDC projects no meaningful consumer memory relief until late 2027–2028. HBM2026 allocations are fully booked; HBM pricing and availability will be a structural constraint on AI capex planning through 2027.

Sources