The Philadelphia Semiconductor Index has climbed above 14,000 points for the first time, marking a record high after surging more than 230% since April 2025. Such a rapid advance has occurred only twice before, last seen during the late 1990s tech boom, underscoring the intensity of the current rally driven by artificial intelligence demand.
Rally driven by memory and advanced manufacturing
Gains have been concentrated in companies tied to memory and high-end chip production. Micron, SK Hynix, and Samsung have risen დაახლოებით 141%, 186%, and 114% this year, while U.S.-listed shares of TSMC are up more than 50%. Nvidia, Broadcom, Marvell, and ASML are trading near or at record levels.
At the center of the surge is a tightening supply of high-bandwidth memory. Goldman Sachs now estimates a 4.9% DRAM supply gap for 2026, up from a previous 3.3% forecast, calling it the most severe shortage in 15 years. Prices reflect that pressure, with HBM3E chips selling for around $300 and next-generation HBM4 potentially reaching $500.
Production capacity is already constrained. SK Hynix’s 2026 output is fully booked, with major clients paying in advance to secure supply. At the same time, AI-driven data center construction continues to expand faster than chip manufacturing capabilities. Only three companies globally can mass-produce high-bandwidth memory, with Hynix controlling roughly half of the market.
TSMC’s advanced CoWoS packaging is another bottleneck, with lead times stretching from 52 to 78 weeks. Nvidia alone has secured an estimated 60–70% of that capacity.
Margins diverge across the supply chain
Not all parts of the semiconductor ecosystem are benefiting equally. Optical module manufacturers are seeing margins shrink despite rising shipment volumes. Chinese firms dominate the 800G and 1.6T markets, but intense price competition is limiting profitability, with further pressure expected as new capacity comes online between 2026 and 2027.
The global semiconductor landscape remains regionally specialized. The United States leads in chip design and AI systems, with Broadcom and Marvell controlling about 95% of custom ASIC orders. Taiwan dominates advanced manufacturing, with TSMC expanding capacity from 35,000 to 130,000 wafers per month by the end of 2026. South Korea leads in memory production, while Japan and the Netherlands maintain critical roles in materials, equipment, and EUV lithography.
Profit structure shifts as manufacturing gains ground
The AI boom is reshaping the industry’s traditional “smile curve,” where design historically captured the highest margins. Advanced manufacturing is now closing that gap. TSMC reported a 66.2% gross margin and 50.5% net margin in early 2026, approaching profitability levels once typical of leading chip designers.
Morgan Stanley expects this trend to broaden. After an initial phase dominated by GPUs, AI-related growth is now spreading into memory, packaging, and networking, pulling new suppliers into the spotlight. ASML, for example, could see earnings per share grow 57% year-on-year by 2027, while companies tied to data center infrastructure such as Schneider Electric, ABB, and Vertiv are reporting stronger-than-expected demand.
Wedbush estimates AI infrastructure spending will reach $725 billion in 2026, a 77% increase from the prior year, while Goldman Sachs projects global AI capital expenditure could rise to $1.6 trillion by 2031.
Volatility and skepticism emerge
Despite strong momentum, the rally has shown signs of fragility. In early June, the index dropped 10% in a single session, erasing more than $1 trillion in market value after a cautious outlook from a major chip designer raised concerns about slowing AI spending.
Some traders are positioning for a reversal. Michael Burry has built large put positions tied to semiconductor ETFs and major AI stocks, drawing comparisons to market conditions before the 2000 tech bubble. Man Group has also warned that AI-related financing structures depend on relatively short-lived hardware assets, with potential stress emerging around 2027–2028.
Earnings and supply outlook in focus
Near-term sentiment will hinge on upcoming corporate results. Micron’s June 24 earnings report is expected to provide insight into whether pricing power in memory remains intact, particularly through its gross margin performance. Nvidia’s next results, scheduled for August 26, are likely to set the broader tone for the sector.
Further out, significant capacity expansions are planned between late 2027 and mid-2028. New facilities from SK Hynix, Samsung, and Micron could increase global supply by 20–30%. However, with demand for HBM growing at more than 40% annually, it remains uncertain whether supply will catch up.
Geopolitical risks remain critical
Structural risks continue to loom over the sector. More than 90% of advanced chip fabrication capacity is concentrated in Taiwan, leaving global supply chains exposed to geopolitical tensions or export restrictions. Any disruption in the region could trigger a rapid reassessment of valuations across semiconductor markets.
The latest surge reflects a market where demand for AI-focused chips continues to outpace supply, concentrating gains among a small group of dominant players. Whether that imbalance can be sustained now depends on earnings performance, capacity expansion, and the durability of AI-driven spending.
Worried this chip rally echoes past bubbles? Learn how sentiment gauges guide rational risk management before volatility hits.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
