Summary

The AI industry is still regarded as the core investment theme for the coming years. However, as valuations for “bottleneck segments” such as GPUs, HBM, and advanced packaging continue to rise, market focus may gradually shift from seeking the scarcest resources to identifying companies with cost advantages and substitution capabilities. Community opinions suggest that future AI investments will require not only assessing industry trends but also balancing valuation levels and technological iteration, with attention to AMD, ASIC, TPU, custom chips, new DRAM suppliers, as well as software and SaaS sectors that have long-term growth potential.

KTX Crypto Portfolio Observations

For the KTX Crypto portfolio, this community perspective represents a thought upgrade on AI industry investment logic, focusing more on possible future capital rotation directions rather than short-term market trends:

• **The long-term AI theme remains unchanged:** AI’s productivity enhancements are still in early stages, and the industry’s long-term growth logic remains valid.
• **Valuation race is more important:** Being optimistic about the industry does not mean chasing the leaders at high prices; it is crucial to combine valuation and market sentiment to find better entry points.
• **Capital may spread from “bottleneck assets” to “alternative solutions”:** As valuations of leading companies rise, firms with cost advantages and substitution abilities may gain more attention.
• **Focus on AI industry chain rotation opportunities:** Beyond hardware like GPUs and HBM, software, SaaS, cloud platforms, and new suppliers could become important beneficiaries in the next phase.
• In portfolio execution, continuously track valuation changes and technological evolution in the AI industry chain, focusing on whether capital is shifting from leaders to alternatives and application layers, while comprehensively judging based on profitability, industry prosperity, and market sentiment.

Original Text Included

The community author, combining recent live broadcasts about AI, U.S. stocks, and the storage industry, summarized several core viewpoints on current AI investments.

The author believes that AI remains the core theme for long-term future development, and its productivity improvements are only just beginning, but the broad industry space does not mean all AI companies are worth chasing at high valuations.

Regarding the industry chain, the storage demand logic for HBM, DRAM, etc., has not changed, and leading companies like SK Hynix and Samsung remain important. However, the current market repricing mainly reflects profit expectations from earlier price increases and pricing power within the industry chain, rather than demand itself.

The author also emphasizes that the macro environment determines the long-term direction, while valuation determines the timing of purchase. Even with a long-term bullish view on AI, a phased accumulation strategy should be adopted to allow room for market fluctuations.

Furthermore, since capital was previously mainly concentrated in hardware areas such as chips, storage, and data centers, future investment opportunities may also arise for application-layer companies that can truly convert AI into commercial revenue, such as software, SaaS, and cloud platforms.

The article presents a noteworthy new viewpoint — AI investment logic may be upgrading from “finding bottlenecks” to “finding cost-effective alternatives.”

In the past, the market focused more on critical supply-constrained links like GPUs, HBM, advanced packaging, and power. But once these areas become market consensus, their valuations often rise significantly, and high profits and costs may drive customers to proactively seek lower-cost alternatives.

For example, although NVIDIA still maintains a leading advantage, AMD, ASIC, TPU, and custom chips are gradually entering the market in different application scenarios; the storage sector may also reduce cost pressures from high-priced storage through long-term procurement agreements, new suppliers, and architectural optimizations. Meanwhile, the development of domestic large models like DeepSeek has made the market realize that AI capability improvements do not necessarily rely entirely on higher computing power investment. Additionally, new DRAM suppliers and other variables may impact future industry supply and pricing.

The author believes that leading companies will not be replaced in the short term, but when researching AI investment opportunities in the future, besides focusing on “who is the scarcest,” it is also worth further considering: who can accomplish the same task at a lower cost.

Original Author: Dr. Moyu | Director Moyu

X Account: @Jason23818126

Original Link: https://x.com/Jason23818126/status/2084295372881727753 

Risk Warning: This article is a collection of community opinions and does not constitute any investment advice. DYOR.