Israel Courts TSMC, Intel, Samsung for New Chip Hub

A Push for Chipmaking Sovereignty

Israel's national AI czar is actively lobbying TSMC, Intel and Samsung to establish advanced semiconductor manufacturing on Israeli soil, aiming to secure a 2-nanometer fabrication facility alongside a data center capable of housing roughly 100,000 GPUs 1. The effort comes as the country grapples with wartime disruptions and constraints on its electrical grid, both of which complicate the pitch to foreign chipmakers who typically demand stable power supplies and long-term security guarantees before committing billions in capital 1. The initiative reflects a broader global scramble among governments to localize chip production and AI compute capacity rather than remain dependent on facilities concentrated in Taiwan, South Korea and the United States.

The Equipment Makers Are Cashing In

While Israel works to attract fabs, the companies that supply the tools for building them are already seeing the benefits of the AI boom. ASML, the Dutch lithography giant whose machines are essential for producing cutting-edge chips at the 2nm class and beyond, posted a second-quarter beat and raised its 2026 outlook, citing surging capital expenditure tied to AI chip demand 4. That upgraded guidance underscores just how much momentum remains behind the buildout of advanced fabs worldwide, and it helps explain why a country like Israel would want a slice of that manufacturing capacity rather than simply importing finished chips.

Nvidia's Dominance Faces New Pressure Points

Nvidia continues to sit at the center of the AI chip conversation, but its position is being tested on multiple fronts. On the military and research side, Nvidia donated a supercomputer built with its latest chips to a nonprofit affiliated with the Naval Postgraduate School, a move that reinforces the company's ties to U.S. government and defense-adjacent institutions even as it fields scrutiny over export controls 5. That scrutiny intensified when a senior White House official accused a Chinese startup of improperly distilling an Anthropic model and accessing restricted Nvidia chips to build a new model called Kimi, a claim that highlights ongoing concerns in Washington about export-control evasion and the leakage of advanced AI compute to China 6.

Robotics Emerges as the Next Battleground

Away from data centers, a parallel contest is unfolding over who will supply the "brains" for humanoid and industrial robots. AMD unveiled its X100 chip lineup, bringing its Strix Halo APU architecture—pairing Zen 5 CPU cores with RDNA 3.5 graphics—into robotics applications designed for continuous, 24/7 operation and a decade-long embedded service life; the chips will also be available through a Kria System-on-Module developer platform aimed at robot builders 2. This is widely read as a direct response to Intel's recent moves into the same space, and it signals that the physical AI market is becoming a genuine battleground beyond traditional data-center GPUs 2.

That competitive pressure on Nvidia is not merely theoretical. Foundation, a robotics company, confirmed that its new MK-2 Phantom humanoid robots will run on AMD Ryzen chips rather than Nvidia hardware, marking a notable win for AMD in a segment where Nvidia has otherwise supplied the vast majority of AI silicon 3. Together, these developments suggest that as humanoid robotics matures from prototype to commercial deployment, chipmakers are positioning early to avoid a repeat of the GPU market's heavy concentration around a single supplier.

Why It All Matters

Taken together, these stories point to an AI hardware landscape being reshaped simultaneously at the geopolitical, industrial and architectural levels. Israel's courtship of TSMC, Intel and Samsung reflects how nations increasingly view chip fabrication as a matter of strategic independence, not just economic opportunity 1. ASML's raised outlook confirms that the capital flowing into AI infrastructure shows no sign of slowing, even as scrutiny over chip access to adversarial states grows sharper, as seen in the Kimi model controversy 46. Meanwhile, Nvidia's continued entrenchment in institutions like the Naval Postgraduate School contrasts with the inroads AMD is making in robotics through both AMD's own product launches and customer wins like Foundation 523. The overall picture is one of an AI chip ecosystem expanding on several fronts at once—manufacturing capacity, government partnerships, export enforcement and entirely new hardware categories like embedded robotics processors—each competing for a central role in how artificial intelligence systems are built and deployed in the years ahead.