Semiconductor Companies Get Tough on AI Investment, but Measuring Returns Remains a Challenge - #semiconductordigest# #semiconductors# #semiconductor# #semiconductorindustry# #semiconductorwebsite# #semiconductormagazine#

SEMICONDUCTORS vs THE S&P 500 This is how both the S&P 500 $VOO and the Semiconductor ETF $SMH have performed so far in 2026 +28.8% - Semiconductors +8.7% - S&P 500 Just a couple weeks ago SMH was up by more than 60% on the year SMH is still outperforming the S&P 500 by more than 3x so far this year If ETFs are interesting for you, you should check out That is where this visual is from 🫡

Semiconductor equipment: The outlook for 2027-2028 continues to improve. A few interesting developments across the semiconductor supply chain following SEMICON West 2026. The overall message is encouraging, particularly for investors exposed to memory, advanced packaging, semiconductor equipment and optical connectivity. Here are my main takeaways: 🟢 1. Visibility into 2027–2028 is improving. Equipment suppliers are receiving customer forecasts much earlier than usual, allowing them to prepare capacity and secure components. Some manufacturers are already planning for a wafer fabrication equipment (WFE) market exceeding $300B in the coming years. This would represent a substantially larger industry than previous cycles. 🟢 2. AI demand continues to broaden. There has been considerable discussion around AI systems requiring less HBM per accelerator, particularly for inference. However, lower memory capacity per individual system does not necessarily mean lower aggregate memory demand. Training remains memory-intensive, while inference is becoming increasingly specialized. Higher deployment volumes, greater bandwidth requirements and additional memory architectures could support continued growth across HBM, conventional DRAM and other memory products. 🟢 3. Advanced packaging is becoming increasingly important. As semiconductor architectures become more complex, manufacturers need additional inspection, metrology, bonding and testing equipment. At 2nm/GAA, process-control intensity could increase approximately 30% compared with equivalent 3nm capacity. This creates opportunities for companies such as $KLAC, $NVMI, $CAMT and $AMAT. Importantly, these businesses can benefit from increasing manufacturing complexity even without equivalent growth in wafer volumes. 🟢 4. Memory technology transitions provide another growth opportunity. Beyond HBM capacity expansion, the industry is preparing for hybrid bonding, 4F² DRAM architectures, taller memory stacks and increasingly complex NAND structures. These transitions require more sophisticated deposition, etch, metrology and process-control equipment. That is encouraging for suppliers such as $LRCX, $AMAT, $KLAC and $MKSI. NAND investment has also remained relatively limited compared with DRAM/HBM, leaving room for additional spending if manufacturers eventually expand capacity. 🟢 5. Silicon photonics is emerging as another significant opportunity. Increasing AI bandwidth requirements are driving interest in silicon photonics, optical interposers and eventually co-packaged optics. Soitec's photonics capacity for 2027–2028 is effectively booked, while $CAMT expects a more meaningful silicon photonics contribution beginning in 2028. $TER is also positioning itself for increasing optical testing requirements. The optical transition is still relatively early, but the investment opportunity extends beyond transceiver and laser manufacturers. 🟢 6. Power requirements continue to increase. AI rack power requirements are moving substantially higher, creating opportunities for precision power conversion and delivery. $AEIS expects another 1–2 generations of higher-power 48V products before the 800V transition becomes meaningful in late 2028 or 2029. 🟢 7. Pricing power is improving across equipment suppliers. More advanced equipment delivers better yields, greater throughput and improved manufacturing efficiency. These capabilities support higher selling prices and potentially better margins. At the same time, some supply-chain components remain constrained, which could limit how quickly manufacturers respond to demand. Key takeaways: The semiconductor investment opportunity is becoming broader than simply forecasting how many GPUs hyperscalers will purchase. Increasing manufacturing complexity, higher memory requirements, more advanced packaging, optical connectivity and power delivery are all creating opportunities across the supply chain. What I find particularly interesting is that several equipment suppliers could benefit from multiple technology transitions simultaneously. The biggest question remains whether demand can sustain the industry's increasingly ambitious capacity plans. But the visibility companies are reporting for 2027–2028 is certainly encouraging.

Semiconductor chokepoints are huge strategic assets: Taiwan is building its Silicon Shield, Korea is near-untouchable due to HBM, America is building AGI with Nvidia chips. Yet despite ASML, Europe captures little AI value and even less strategic leverage. Why? (1) No BATNA. If ASML stops exporting to Taiwan, it loses all its revenue, if it starts exporting to China, America starts cutting off essential components. And it sells no machines to Europe at all--so whenever anyone starts thinking about 'using' the ASML chokepoint, everyone else calls the bluff. (2) No political awareness. European governments are trapped in dogmatic slumber, unwilling and unable to treat ASML as a strategic asset. As a result, Washington feels like it has to run European export policy instead. Every other country uses its bottlenecks, but Europe pretends it can’t. (3) No inspiration. There's no project, no vehicle, nothing to actively choose. Weaponising export controls just to annoy great powers isn't very motivating; policymakers ask 'what for?'. Right now, there's no good answer. That means there's a gap in Europe's AI strategy: an ambitious play, a conduit for political attention, and an alternative use for ASML machines all at once. The timing is right: all this is happening just as the chip industry is about to be upended and accelerated by AI systems that uplift chip and process design. In the next years, we might close the hardware loop for recursive self-improvement: use models to build better chips which we use to train even better models to build even better chips. For that loop, chip fabrication is one central bottleneck. That's why I believe the gap in Europe's strategy is fab-shaped. I suspect the best way to fill it is to enlist and subsidise ASML and imec and to recruit TSMC and American frontier labs to build a leading R&D chip facility in Europe: Eurofab. from this week’s essay:

Semiconductor prices are skyrocketing in the US: Producer prices for semiconductor and other electronic component manufacturing surged +26% YoY in September, their 4th-largest increase on record. This also marks their 6th consecutive YoY increase of at least +25%. Since July 2025, the Producer Price Index for semiconductor and other electronic component manufacturing has risen +31%, to its 2nd-highest level since January 2007. To put this into perspective, the index declined every year from 1994 to 2021. Meanwhile, producer prices for electronic components and accessories soared +27% YoY last month, also marking their 6th consecutive YoY increase of at least +25%. The AI Revolution is driving historic technology inflation.