门捷列夫学徒
0 Following
42 Followers
25 Liked
1 Shared
See original
The weekend lithium mines and anti-involution are so hot~
However, I am paying attention to another anti-involution sector (express delivery), I think the chips might be better, let's see if it can have a performance like the new consumption in May (Ru Yuchen, Wancheng, Pets).
Additionally, from the perspective of chips, several sectors in August looked too much like blood bags
[cover face]
However, I am paying attention to another anti-involution sector (express delivery), I think the chips might be better, let's see if it can have a performance like the new consumption in May (Ru Yuchen, Wancheng, Pets).
Additionally, from the perspective of chips, several sectors in August looked too much like blood bags
[cover face]
See original
Some thoughts:
Some thoughts:
1. In this round of bull market, I see many people guessing the top. In fact, 4000, 5000, and 6000 are meaningless, and the key is to look at the real estate market. Every past bull market has been accompanied by a jump in housing prices, a large transfer of funds, and a drain on funds. If this round of the real estate market is still like this, then the bull market will have no top, reshaping a generation's asset beliefs. Conversely, if not, then run quickly, history is still repeating itself.
2. Trump has indeed done a great job this time. The EU, Japan, and South Korea have all kneeled, and funds have flowed back to the United States in large amounts, which is good for Nasdaq and AI infrastructure. In fact, to analyze various assets, you have to look at the flow of money.
1. In this round of bull market, I see many people guessing the top. In fact, 4000, 5000, and 6000 are meaningless, and the key is to look at the real estate market. Every past bull market has been accompanied by a jump in housing prices, a large transfer of funds, and a drain on funds. If this round of the real estate market is still like this, then the bull market will have no top, reshaping a generation's asset beliefs. Conversely, if not, then run quickly, history is still repeating itself.
2. Trump has indeed done a great job this time. The EU, Japan, and South Korea have all kneeled, and funds have flowed back to the United States in large amounts, which is good for Nasdaq and AI infrastructure. In fact, to analyze various assets, you have to look at the flow of money.
See original
A couple of points before the market opens, regarding applications and domestic computing power:
1. The overall big beta for AI applications in the A-shares market should be gone (note that the premise is A-shares). Currently, the mainstream view in North America is that the opportunities for agents belong to the giants, as only the giants can invest in computing power. The opportunities for small companies in the future may resemble the MCNs around Douyin that focus on high-quality content and creativity. If you believe that software companies in the A-shares market have opportunities, it also means that those small software outsourcing companies in the office building next door will experience a qualitative change (which is obviously impossible).
2. The H20 matter was initially thought to be insignificant, but the more we look at it, the more it seems not to be quickly resolvable; let's wait for a turning point. Or when can domestic computing power break through? However, this big beta for domestic computing power should eventually come. In the short term, we still need to wait.
The clear opportunities are still in the North American chain, with the main focus on optical modules and PCBs. Besides these, there have been quite a few new directions and breakthroughs in sectors recently.
1. The overall big beta for AI applications in the A-shares market should be gone (note that the premise is A-shares). Currently, the mainstream view in North America is that the opportunities for agents belong to the giants, as only the giants can invest in computing power. The opportunities for small companies in the future may resemble the MCNs around Douyin that focus on high-quality content and creativity. If you believe that software companies in the A-shares market have opportunities, it also means that those small software outsourcing companies in the office building next door will experience a qualitative change (which is obviously impossible).
2. The H20 matter was initially thought to be insignificant, but the more we look at it, the more it seems not to be quickly resolvable; let's wait for a turning point. Or when can domestic computing power break through? However, this big beta for domestic computing power should eventually come. In the short term, we still need to wait.
The clear opportunities are still in the North American chain, with the main focus on optical modules and PCBs. Besides these, there have been quite a few new directions and breakthroughs in sectors recently.
See original
Seeing Boss Sh discussing the issue of valuation, institutions are certainly concerned about the 'new energy' transformation of AI. I would like to mention two points:
1. New energy faces the issue of single product penetration rate, while for AI, it is difficult to measure penetration rate. Computational power is more like electricity, which is a fundamental facility. You can at most measure the penetration rate of 'light bulbs', but electricity demand has been continuously growing for a century. Today, we are still investing in electricity (Yarlung Tsangpo River), so we cannot apply the new energy paradigm here.
2. Computational power will not always grow exponentially; if the growth rate declines, will it also experience internal competition and losses like new energy? The answer is no, because the price of new energy is truly determined by the upstream, whether it is photovoltaic or lithium battery, the upstream is bulk commodities, which are particularly sensitive to marginal changes. However, there is no such upstream in the computational power chain, and the competitive landscape is stable. Note that not every industry experiencing a slowdown in growth will necessarily lead to a price war; only those that are particularly sensitive to upstream changes will (otherwise, why hasn't Apple lowered its prices?).
In summary, I firmly believe that the computational power chain has room for valuation improvement. As for how large domestic funds with pricing power think, that is their issue. $Zhongji Xuchuang (SZ300308)$ $Shenghong Technology (SZ300476)$ $Xinyi Sheng (SZ300502)$
1. New energy faces the issue of single product penetration rate, while for AI, it is difficult to measure penetration rate. Computational power is more like electricity, which is a fundamental facility. You can at most measure the penetration rate of 'light bulbs', but electricity demand has been continuously growing for a century. Today, we are still investing in electricity (Yarlung Tsangpo River), so we cannot apply the new energy paradigm here.
2. Computational power will not always grow exponentially; if the growth rate declines, will it also experience internal competition and losses like new energy? The answer is no, because the price of new energy is truly determined by the upstream, whether it is photovoltaic or lithium battery, the upstream is bulk commodities, which are particularly sensitive to marginal changes. However, there is no such upstream in the computational power chain, and the competitive landscape is stable. Note that not every industry experiencing a slowdown in growth will necessarily lead to a price war; only those that are particularly sensitive to upstream changes will (otherwise, why hasn't Apple lowered its prices?).
In summary, I firmly believe that the computational power chain has room for valuation improvement. As for how large domestic funds with pricing power think, that is their issue. $Zhongji Xuchuang (SZ300308)$ $Shenghong Technology (SZ300476)$ $Xinyi Sheng (SZ300502)$
See original
Just chatting casually over the weekend.
Just chatting casually over the weekend.
First, let's talk about AI. Since GPT-4 in 2023, there have been two major 'negative factors' surrounding AI: first, hardware is not shipping (tariffs, production scheduling), and second, AI intelligence has reached a training bottleneck. This year, in the hot summer of AI in 2025, H20 (hardware) and the report on GPT-5 by The Information over the weekend may not meet expectations, and coupled with the fact that the sector is indeed at a high level, different voices are emerging.
What I want to say is that in the past two years, AI has been disturbed by countless such voices (this time it is really very lightweight), but each time has proven to be interference and noise, which does not affect NV and its partners from continuing to reach new highs; this is the power of the trend. Recently, I watched the latest interview with the two co-founders of Anthropic, Dario and Mann (by the way, Claude code is really an AI application product I am very optimistic about). When the big shots mentioned the scaling law, I felt it is more verifiable than many physical laws. Mann's original words were that there are very few such laws in the universe that still hold when the scale expands by several hundred thousand times (look at quantum mechanics and general relativity
First, let's talk about AI. Since GPT-4 in 2023, there have been two major 'negative factors' surrounding AI: first, hardware is not shipping (tariffs, production scheduling), and second, AI intelligence has reached a training bottleneck. This year, in the hot summer of AI in 2025, H20 (hardware) and the report on GPT-5 by The Information over the weekend may not meet expectations, and coupled with the fact that the sector is indeed at a high level, different voices are emerging.
What I want to say is that in the past two years, AI has been disturbed by countless such voices (this time it is really very lightweight), but each time has proven to be interference and noise, which does not affect NV and its partners from continuing to reach new highs; this is the power of the trend. Recently, I watched the latest interview with the two co-founders of Anthropic, Dario and Mann (by the way, Claude code is really an AI application product I am very optimistic about). When the big shots mentioned the scaling law, I felt it is more verifiable than many physical laws. Mann's original words were that there are very few such laws in the universe that still hold when the scale expands by several hundred thousand times (look at quantum mechanics and general relativity
See original
Another new high this year. When a sector is doing well, everything will thrive, so those sect disputes are so boring (and many people who barely pass their research dare to write lengthy articles).
Additionally, AI has definitely not reached its end; this is just a small climax.
Additionally, AI has definitely not reached its end; this is just a small climax.
See original
This market is indeed too good, and it seems there is still a lot of incremental capital.
See original
With the deepening of the anti-involution, an ancient and long-dormant powerful sector is about to be awakened, which is the coal dance. A large amount of old money lurking outside is already eager to move, and the market trading volume is unlikely to drop below 1.5 trillion.
After all, whether it’s AI or innovative drugs, most of the people involved are still new money. I personally have a lot of respect for old money; its durability and explosiveness are very much worth looking forward to. Everyone has their own comfort zone where they excel. After all, making money is the hard truth, and just copying the homework of predecessors is enough. $Tianqi Lithium (SZ002466)$ $Northern Rare Earth (SH600111)$
Of course, I still primarily focus on the growth sectors represented by AI, as growth and cycles are always the best themes in A-shares.
After all, whether it’s AI or innovative drugs, most of the people involved are still new money. I personally have a lot of respect for old money; its durability and explosiveness are very much worth looking forward to. Everyone has their own comfort zone where they excel. After all, making money is the hard truth, and just copying the homework of predecessors is enough. $Tianqi Lithium (SZ002466)$ $Northern Rare Earth (SH600111)$
Of course, I still primarily focus on the growth sectors represented by AI, as growth and cycles are always the best themes in A-shares.
See original
MPO is too awesome
[Dog Head]
$Tai Chen Guang (SZ300570)$ $Chang Xin Bo Chuang (SZ300548)$ $Shi Jia Guang Zi (SH688313)$
[Dog Head]
$Tai Chen Guang (SZ300570)$ $Chang Xin Bo Chuang (SZ300548)$ $Shi Jia Guang Zi (SH688313)$
See original
I hope everyone patiently takes care of the slow bull; the key is to nurture the 'slow' and not worry about the bull
[Dog Head]
Additionally, I think there should be another big opportunity for domestic AI or applications within this year, and I mean the kind of opportunity that doubles; of course, North America hasn't finished yet.
[Dog Head]
Additionally, I think there should be another big opportunity for domestic AI or applications within this year, and I mean the kind of opportunity that doubles; of course, North America hasn't finished yet.
See original
Weekend Talk: Foundations of a Bull Market, AI, and Trading
Long-term funds entering the market are certainly a necessary condition for a bull market, but this is only a result and a surface phenomenon. The core is whether there are enough high-quality growth assets in the market. Note that it is growth, not cheap assets.
Yushu is about to go public, and there’s also Zhiyuan Robotics, which is borrowing the shell of Weiqi New Materials, plus I’ve learned that many large models (like Zhipu) are also seeking to go public. The market seems to have controversy over this, questioning why humanoid robots or large models are in a hurry to go public without large-scale commercialization. Is it a rush to cash out? I believe that allowing a large number of emerging industries represented by Yushu to go public early is precisely the core foundation of a long-term bull market. It allows more retail investors to share in the fruits of leading growth in emerging sectors.
Yushu is about to go public, and there’s also Zhiyuan Robotics, which is borrowing the shell of Weiqi New Materials, plus I’ve learned that many large models (like Zhipu) are also seeking to go public. The market seems to have controversy over this, questioning why humanoid robots or large models are in a hurry to go public without large-scale commercialization. Is it a rush to cash out? I believe that allowing a large number of emerging industries represented by Yushu to go public early is precisely the core foundation of a long-term bull market. It allows more retail investors to share in the fruits of leading growth in emerging sectors.
See original
North American chain is really exciting, it is estimated that it will force a short squeeze and rise further
See original
I don't want to write a long article anymore. Those who are paying attention to AI applications now should see who takes the high ground of AI coding. Recently, both Silicon Valley and domestic ByteDance have focused on this. AI is changing very quickly; the grand narratives of previous AI software are temporarily unimportant. AI coding is the key to short-term applications or the realization of ASI. Recently, the computing power flywheel has also shown everyone the opportunity for deterministic AI coding. Coding is the closed loop of ASI in the digital world, and it is very easy to verify and iterate. Next comes Agent empowering humanity, and the ultimate goal is physical AI. This time, ByteDance has taken the most and the firmest stand, strategically pushing hard on coding. Personally, I believe ByteDance is still the company most closely aligned with Silicon Valley. ByteDance's related computing power is the first to release performance (T may be the slowest, A in the middle). Therefore, those so-called AI applications from AH need to distinguish which ones are catching up; in fact, computing power is the most certain.
See original
Said yesterday not to reduce positions
[Dog Head]
Today can reduce, but the overall trend is still upward.
Previously mentioned in the posts that some domestic computing power can also release performance in the second half of the year, for example, Chai Fa has been three consecutive gains, and I personally have high hopes for switches, which have larger flexibility. Additionally, there are some niche questions that many people are asking about, but it is estimated that the release won't be that fast.
[Dog Head]
Today can reduce, but the overall trend is still upward.
Previously mentioned in the posts that some domestic computing power can also release performance in the second half of the year, for example, Chai Fa has been three consecutive gains, and I personally have high hopes for switches, which have larger flexibility. Additionally, there are some niche questions that many people are asking about, but it is estimated that the release won't be that fast.
See original
Apart from Chai Fa, the mainstay of domestic computing power has changed its face, such as the server companies that are OEMs for B30 and domestic switch manufacturers. Especially for switch companies, although the current price-to-earnings ratio is high, the future potential should be significant. Changing the face is quite good; the structure of chips for computing power leasing is indeed not favorable inside. This wave of domestic computing power has temporarily not participated in leasing.
After choosing the Taiwanese market leader, the Agent market has a higher outlook.
North American computing power optical interconnection is still very optimistic, and XC hasn't reached a new high yet.
After choosing the Taiwanese market leader, the Agent market has a higher outlook.
North American computing power optical interconnection is still very optimistic, and XC hasn't reached a new high yet.
See original
To say something unpleasant, those holding optical modules/copper cables/CPO are arguing without even understanding the basic concepts. Clearly, these are all symbiotic relationships, used to distinguish between in-cabinet and out-of-cabinet. If you want to say that there were indeed technical route disputes and replacement relationships in the past regarding photovoltaics, that would be one thing, but these few in AI are just about the rhythm of performance release and differences among downstream customers.
See original
Grok4 is really impressive, a product of a 200,000-card cluster, pre-training is truly miraculous, and the scaling law is still effective.
Soon Xai will use 100,000 cards GB200 to train multimodal agents, expected to be released around October this year, with future plans for a million-card cluster.
DS R1 and the most advanced models in North America are no longer from the same generation. Remember when DS first came out this year, many people were still saying that computing power wasn't important, such anti-intellectual remarks. Now there are also voices saying that it's right for CSPs like Alibaba to focus on delivery, putting AI on hold, which is really frustrating~
Soon Xai will use 100,000 cards GB200 to train multimodal agents, expected to be released around October this year, with future plans for a million-card cluster.
DS R1 and the most advanced models in North America are no longer from the same generation. Remember when DS first came out this year, many people were still saying that computing power wasn't important, such anti-intellectual remarks. Now there are also voices saying that it's right for CSPs like Alibaba to focus on delivery, putting AI on hold, which is really frustrating~
See original
MPO has been very strong these days, and I also think that copper and AEC in the scale-up direction are relatively undervalued (today's strong BC is not without reason). Recently, the increase has lagged significantly behind the scale-out optical modules, which are also extremely cheap. $Changxin Bochuang(SZ300548)$ $Taichengguang(SZ300570)$
Recently, North American investment bank reports are very straightforward, just continuously raising, raising, and still raising.
Recently, North American investment bank reports are very straightforward, just continuously raising, raising, and still raising.
See original
Are these trillion-dollar internet giants in the country really not considering the impact on the capital market and shareholder rights when making decisions? Why does it still feel like the strong localism of a small shop at the village entrance?
Even if you win the battle for food delivery, the increase in Meituan's market value only translates to a 40% impact on Alibaba. Meanwhile, the seven sisters in the US stock market have increased several times due to AI.
Now, if the future continues to be mired in the food delivery war, it's possible that this time, the combined decline of several internet giants could equal the market value drop of Meituan. $Alibaba (BABA) $
Even if you win the battle for food delivery, the increase in Meituan's market value only translates to a 40% impact on Alibaba. Meanwhile, the seven sisters in the US stock market have increased several times due to AI.
Now, if the future continues to be mired in the food delivery war, it's possible that this time, the combined decline of several internet giants could equal the market value drop of Meituan. $Alibaba (BABA) $
See original
In several AI segments (scrap), perhaps only a few upstream industries like CCL and electronic fabric will truly benefit from this wave of AI (but their profit release potential is far less than that of optical modules back in the day). The rest are mostly just thematic hype, but it doesn't affect stock price increases. The A-share market loves a new story, and everyone gets what they need. Unfortunately, several prominent targets have very large market capitalizations already; if they rise further, the calculations will become too easy.
[Dog Head]
: The market capitalization is almost exceeding that of the authentic AI chain like MPO, AEC, and copper connectors, and the valuation is simply not worth looking at.
Other random thoughts: Besides North American AI, I think the gaming sector has very good chips, and there are also some solid chips related to stablecoins in cross-border payments. The ones with exhausted chips are new consumption and robotics. Additionally, the Hang Seng Index has adjustment needs due to the delivery battle (I don't really believe that innovative drugs can support the Hang Seng Index on their own; there have been obvious signs of distribution recently), and the Shanghai Composite Index has also reached 3500, which exceeds the psychological level. Overall trading opportunities in July might not be as good as last month.
[Dog Head]
: The market capitalization is almost exceeding that of the authentic AI chain like MPO, AEC, and copper connectors, and the valuation is simply not worth looking at.
Other random thoughts: Besides North American AI, I think the gaming sector has very good chips, and there are also some solid chips related to stablecoins in cross-border payments. The ones with exhausted chips are new consumption and robotics. Additionally, the Hang Seng Index has adjustment needs due to the delivery battle (I don't really believe that innovative drugs can support the Hang Seng Index on their own; there have been obvious signs of distribution recently), and the Shanghai Composite Index has also reached 3500, which exceeds the psychological level. Overall trading opportunities in July might not be as good as last month.
Login to explore more contents