This weekly newsletter pulls together summaries of the top ten most-read Insights across Tech Hardware and Semiconductor on Smartkarma.
Receive this weekly newsletter keeping 45k+ investors in the loop
1. Silicon Wafers & The Looming Leading Edge Supply Crisis
- We are upgrading our existing facilities and any discussions with customers at this point are based on the capacity that we believe is possible as a result of such upgrades.
- Demand is likely to exceed supply for more than three years, because of such supply constraints. In terms of the price level, prices would need to be dramatically higher.
- We have no plans whatsoever to restart the Yoshinogari project. At current price levels, it is not viable. All costs are rising now, with Japan potentially headed for more inflation
2. AMD @ Goldman Sachs Communacopia + Technology Conference 2026. Key Takeaways & Full Transcript
- AMD raised its 2030 Server CPU Total Addressable Market (TAM) projection from $60 billion to $220 billion, targeting >50% market share as it scales multi-threaded architectures like “Venice”.
- Shifting from isolated CPU/GPU/FPGA product roadmaps to a combined Data Center Roadmap Steering Group. Hardware and software trade-offs are now evaluated holistically to deliver full rack-scale system solutions like Helios.
- AMD targets a >60% CAGR in Data Center revenue and >80% CAGR in AI Data Center revenue over its strategic timeframe, driving overall company growth above 35% annually.
3. NVIDIA @ Goldman Sachs Communacopia + Technology Conference 2026. Key Takeaways & Full Transcript
- NVIDIA reaffirmed its confidence in delivering 70% year-over-year revenue growth for the coming year, noting unconstrained demand exceeds 100%.
- While software coding remains primary high-ROI use case, Huang highlighted cybersecurity as the next massive application, with NVIDIA deploying swarms of Nemotron models for red-teaming (bug finding) and blue-teaming (patching)
- NVIDIA is actively working with the financial sector to position its high-utility GPU platforms as durable, asset-backed collateral for institutional financing. Seem like the strategy is working well, for now…
4. Jensen Huang Chats With President Trump @ The All In Summit. Key Takeaways & Full Transcript
- Frontier AI Labs: Pausing, pacing, those are all the voluntary things that they could do if they feel that their company is out of control.
- President Trump: And I’m telling you, it’s all a hoax. The data centers are great and they make people wealthy and they make states wealthy
- Jensen Huang: I’d rather help everybody succeed instead of taking a slice out. And so we would go up as far as we need to but as low as possible.
5. Satya Nadella @ The All In Summit. Key Takeaways & Full Transcript
- Nadella argues the industry must address AI risks through classic, rigorous engineering practices: transparent chain-of-thought (CoT) tracking, strict behavioural auditing and independent, third-party testing rather than “cozy arrangements”.
- Nadella wants broad industry interoperability standards (e.g., externalising memory/harnesses from specific models and creating unified KV cache standards across models) so customers can switch between models without losing their data.
- While continuing its deep strategic partnership with OpenAI, Microsoft is simultaneously building its own native AI models and adopting heterogeneous chip infrastructure (NVIDIA, AMD, internal silicon, and OpenAI’s custom silicon)
6. TechChain Insights Q&A: Kinik Automating Its Plants to Add Advanced-Node Capacity Ahead of Schedule
- Kinik says its biggest challenge is accelerating factory expansion and automating production lines against a labor shortage, so it can deliver capacity ahead of schedule into high demand.
- The capacity plan is dated: 12 inch wafer capacity goes from 330,000 pieces a month to 400,000 by the third quarter, all of the addition high grade reclaim.
- Concentration at the leading node helps a high end supplier: gross margin reached 37.6% in the first half of 2026 against 32.5% for 2025 and 31.3% for 2024.
