While U.S. suppliers are getting 15% more orders for early-2026 parts – mostly AI chips and sensor modules – hardware engineering took 30% of Apple’s new hires this month, all brought in to drive the next round of devices.
Ternus’s teams are building the next Macs and Vision products, coordinating the U.S. factory expansion, and pushing everything Apple needs finished before next year’s launches. And even before the board makes its final call on Cook’s successor, Ternus was seen in Apple’s Regent Street store in London, a role Cook usually had at the Fifth Avenue location.
Apple also updated its financial report this week, confirming that research costs reached their highest level in more than two years during a busy quarter for its main tech work.
AI Training and Silicon Labs Take the Biggest Share of Current R&D Spending
Apple spent $34.55 billion on research this year, 10% more than in 2024, and doubled its U.S. R&D spending over the past five years, with AI infrastructure and silicon engineering as top priorities. Operating costs for AI research jumped 11% year-over-year, with projections pointing to a 30% spike in 2027.
At the same time, the real difference between Apple, Amazon, Meta, or Microsoft won’t be only a huge data center with Nvidia chips costing tens of thousands of dollars each. Apple runs its AI processing on the same chips that power MacBooks and Macs – hundreds of dollars per chip instead.
AI spending will surpass $500B next year – Meta’s $72B, Microsoft’s $80B, Alphabet’s $75B, Amazon’s $100B – and such demand pushed companies to raise compute and storage investment by 166%, reaching $82B in Q3.
Outside tech, financial services are pouring billions into the infrastructure, racing to integrate AI-powered fraud detection and real-time payment processing as online payments jumped from $6.3 billion in 2017 to over $95 billion in 2024.
But despite those numbers, Wall Street money started flowing into crypto only when BlackRock and Abrdn tokenized money market funds in late 2024. Initial Coin Offerings pulled in $6.5 billion in 2022, and 2026 looks different – institutions now represent 19% of ICO investors, up from nearly zero three years ago.
North America hosts 26.4% of launches, Asia-Pacific 24.9%, with over 2,000 ICOs going live through AI infrastructure, healthcare platforms, and decentralized finance. With new tech, investors can use KYC/AML compliance, security audits, and team credentials to see if the current best ICOs to invest in actually have the required features and experienced teams behind them.
For many retail investors, it became part of online culture when early Solana investors paid $0.22 in 2018 and watched it at $260.06 – nearly 12,000% gains. Ethereum’s $18.4 million raise created a crypto worth over $200 billion today.
Both projects needed years of infrastructure development and chip manufacturing capacity to scale, the same challenge Apple faces domestically.
Apple’s Engineering Network Leads through 2026 as Supply and Silicon Units Hit Full Speed
Ternus’s hardware teams are building the next Mac lineup and Vision products while coordinating the entire U.S. manufacturing expansion – most projects Apple wants shipped before next year’s cycle.
GlobalWafers America in Sherman, Texas, makes advanced 300mm wafers using American silicon from Corning’s Hemlock Semiconductor. These wafers flow to American chip fabs such as TSMC in Phoenix and Texas Instruments in Sherman, where they’re manufactured into chips for iPhone and iPad devices sold globally.
Texas Instruments operates facilities in Lehi, Utah, and Sherman, and the partnership grows significantly under Apple’s new commitments, with additional tool installations supporting production of semiconductors.
Amkor’s Arizona facility will handle the final step in manufacturing silicon chips — packaging and testing — as part of a new $7 billion campus planned to support U.S. production. Apple is expected to be one of the first major customers once the site opens, alongside companies like Nvidia. The facility is being built to package Apple silicon manufactured at the nearby TSMC fab, so chips will no longer need to be shipped back to Taiwan for packaging, which was the problem until this facility came online.
Samsung’s Austin facility represents another piece of Apple’s domestic manufacturing puzzle. The two companies are launching chip-making tech that’s not widely used anywhere in the world, with Samsung’s fab supplying chips that optimize power and performance for Apple products.
Apple is also working with Broadcom and GlobalFoundries to develop and manufacture additional cellular semiconductor components in the U.S. – crucial for 5G communications in Apple products.
The Houston server factory produced its first test unit in July 2025 – a 250,000-square-foot facility is set to begin mass production in 2026, creating thousands of jobs according to Apple’s projections. These servers were previously manufactured outside the U.S., but they’ll play an important role in powering Apple Intelligence and serve as the foundation of Private Cloud Compute.
The servers deliver the security and performance of Apple silicon to the data center, and Apple’s teams designed them to be incredibly energy efficient, reducing the energy demands of Apple data centers that already run on 100% renewable energy.
