
Rocky_BTC
Rocky_BTC
Long term investor #BTC #TAO #SOL #SUI #XRP #OKB| MeMe Professional Data Player | Crypto since 2017 | Not financial advice, DYOR🙏Twitter:@Rocky_Bitcoin
5Following
210followers
Feed
Feed
Today I read a UBS analysis report on Broadcom $AVGO and was shocked. According to UBS's forecast, by 2028 Broadcom will become a company worth at least 5 trillion USD! 🧐
Honestly, this projection is quite large. Let me break it down and calculate for everyone to see how much real money $AVGO can make from shipping 35 million TPU units by 2028! 🧮
UBS provided TPU market data in the report: 35 million TPU units, with Broadcom conservatively estimated to hold 60% market share, about 20 million devices, each priced at 15,000 USD, conservatively estimating 300 billion USD TPU revenue.
Additionally, with the surge in ASICs, the market demand for AI network infrastructure will explode, driving up demand for Tomahawk 6/7 network switches and PCIe switches. Network communication is expected to contribute about 40 billion USD.
Traditional business + VMware software: assuming steady or stable annual growth, contributing about 35 billion USD.
📌 Total revenue here will astonishingly reach 375 billion USD, which is 5 times the 2025 level.
Based on this year's Q1 net profit margin of 42%, the estimated net profit for 2028 will reach an astonishing 157.5 billion USD. To put this in perspective, Nvidia's full-year net profit for fiscal 2026 is only 120 billion USD.
Using Nvidia's current PE ratio of 41x to value Broadcom, the market cap would reach an astonishing 6.45 trillion USD. Even with a conservative 30x PE, it would be 4.7 trillion USD, 2.6 times the current market cap! 🧐
But I still feel something is off. This projection is too big and could choke. Let me point out some major potential pitfalls here 🚨
1️⃣ Supply chain bottlenecks:
TPU shipments growing from 4.3 million to 35 million in two years is an 8x increase, which will severely test supply chain capacity. TSMC's capacity, CoWoS packaging technology, HBM memory supply (Micron, SK Hynix, Samsung) — every link must keep up. Currently, Nvidia's GB300 and VR200 are competing for capacity. Can AVGO secure enough wafer allocation? This is a big question.
2️⃣ Customer concentration risk:
AVGO's ASIC main customers are just a few:
• Google (TPU)
• Meta (MTIA)
• OpenAI (rumored to be developing, cooperating on Nexus 10GW project)
• Anthropic (3.5GW agreement)
If any of these major customers cut or delay orders, AVGO's ASIC business will collapse immediately. Moreover, these tech giants are burning huge amounts on AI; if the economy downturns and capital expenditures are cut, ASIC orders will shrink significantly.
3️⃣ Intensifying competition:
ARM AI ASIC chips are also on the way, expected to launch in early 2027, and MediaTek is working on dual sourcing. Although the market pie is growing, more players are sharing it. Whether AVGO can maintain its market share is a key variable.
📝 My personal judgment is that UBS's report looks great on paper, but the actual implementation has too much uncertainty. I think 35 million shipments is somewhat unrealistic; a more realistic expectation might be around 15 to 20 million units.
If we assume 15 million units with 60% market share, AVGO would get 9 million units, and the market cap would double by 2028. That should be quite achievable.
Key time points to watch:
• End of 2026 to early 2027: ARM AI ASIC launch, watch the competitive landscape.
• Mid-2027: Monitor OpenAI's ASIC project progress, which will greatly impact AVGO.
• 2028: See if actual shipments can meet UBS's expectations.
DYOR 🙏, investing carries risks. Just take this UBS report as a reference and don't take it too seriously!




South Korea🇰🇷 stock market had a brief flash of recovery during the session, but ended up crashing again!
The Korea Exchange got anxious and started strict monitoring and investigation of illegal short selling!
Now, this huge pressure is hitting tonight's US stock market😭
Calling📞 Trump, Jensen Huang, Musk…

Facing different markets, Chinese and American robots have taken completely different approaches! 🧐
🇨🇳 Chinese robots focus on emotional value, rescuing the single-person economy!
Core pain points: loneliness, empty nests, and high-pressure urban life.
🇺🇸 American robots focus on hardcore manual labor, tackling blue-collar costs head-on!
Pain points: high blue-collar labor costs, difficulty hiring, and efficiency bottlenecks.
👈 is UBTECH 👉 is Boston Dynamics
Ultimately, this benefits the robotics sector. This topic has been gaining momentum recently, with related keywords ranking top three on Twitter, Xiaohongshu, and Reddit. You might want to keep an eye on related concept stocks! 🧐
Rocky_BTC
The three major technological revolutions of this era:
1. AI
2. Robotics
3. Quantum Computing
With AI being overheated in speculation, it’s worth studying the latter two!
Currently, the robotics sector hasn’t reached its GPT moment yet, and I believe its importance is on par with AI.
I often read @citrini’s memos, which can be said to be the most insightful research institution in the tech world, especially the dystopian scenario report "Autonomous AI Agent Out-of-Control Spread and the 2028 Global Intelligence Crisis," which directly dragged down SaaS software stocks at the beginning of the year—undeniable strength!
Today I saw Citrini invested in Robostrategy, which is an important signal. If you don’t know which robotics company to invest in, you might as well take a look at Robostrategy, stock code #BOT. It is an investment fund focused on the robotics sector and surged 11.8% against the market trend yesterday.
It has invested in many promising and unlisted robotics unicorn🦄 companies’ equity, such as Figure AI, Apptronik, Standard Bots, etc. Worth keeping an eye on🧐


In China, these high-dividend blue-chip assets have once again reached a historic high!
In mainland China, most disposable assets are held by people aged 70-80, and their favorite are high-dividend blue-chip assets, providing steady happiness! 🧐

Rocky_BTC
2.1 million in cash, using the classic 631 rule: 60% for buying stocks, 30% for bonds, and 10% in cash!
Stocks: 1.26 million, buy China Shenhua, China Merchants Bank, or Industrial and Commercial Bank of China; just pick one bank, as their business is stable, dividends are consistent, and the dividend yield is greater than 5%. Calculating at 5%, that’s 63,000 per year!
Bonds: 630,000, directly buy low-volatility fixed income products through Alipay. It is recommended to prioritize China Merchants, Guangfa, Ping An, and E Fund, which generally yield about 3%-5% annually. Calculating a compromise at 4%, that’s 25,200!
Cash: 210,000, for personal living expenses or emergency funds for family illnesses. At 40 years old, middle-aged, with aging parents and young children, there may be times when money is needed!
Estimated total annual income is 88,200, about 7,350 per month. If things go well, with positive stock returns + dividends, it could exceed 100,000! You can also take on a regular job or drive a Honda Accord for Didi + ride-sharing, earning 10,000 a month, living comfortably in a 2-3 tier city 😌!


🚨Why do I get hit with explosive news about Micron every day?! 🥹
Rumor has it that Google $GOOG released an AI tool called TurboVec that can reduce memory requirements by 92%!
But I checked the paper, and this is clearly last year's paper being hyped up again? The short sellers are really nasty!
Why do these ghost stories happen several times every year! 🧐
I suggest everyone feed the paper to AI; it actually has no impact on HBM at all!


Leonard Rodman
SAM ALTMAN HAS A NEW PROBLEM. 🤯
Google just shrunk 31GB of AI memory down to 4GB.
The tool is called TurboVec.
It uses up to 16x less memory, searches faster than FAISS, runs fully offline, and works on a regular Mac.
No expensive GPU cluster.
No cloud dependency.
No compromise on speed.
→ 16x lower memory usage
→ Faster vector search
→ Works with LangChain & LlamaIndex
→ 100% open source
The race to build bigger AI models is loud.
The race to make them dramatically cheaper just got a lot more interesting.
Repo:


After returning from Huaqiangbei in Shenzhen this time, I saw my good brother hoarding MLCC with real money. He told me that this thing is currently in short supply, and those who have stock now are the real winners; the prices for procurement are rising!
Today, seeing the whole internet discussing storage, I spent a day researching MLCC 🧐
I decided that with this wave of decline 📉, I will gradually increase my position in Murata (#MARRY), which is the absolute leader in the global AI server high-capacity MLCC market, holding a 45% market share and possessing very strong pricing power. On April 1st this year, it was the first to raise prices by 15%-35%, showing absolute dominance!
Especially their flagship product, the 0402 inch ultra-thin, ultra-high-capacity MLCC, which is currently a globally exclusive product. This extreme miniaturization perfectly fits HBM3e / HBM4 and the next-generation GPU Nvidia Rubin architecture, capable of solving high-frequency decoupling and power supply stability issues for high-power GPUs and HBM in a very small physical space. This should be considered a core fortress in the era of high-density computing power. Even the first-tier competitors like Samsung Electro-Mechanics and Taiyo Yuden are still chasing and unable to achieve mass production!
Additionally, from the latest financial report data 📊, this might be one of the few AI hardware companies I am willing to invest in right now. After this pullback, its forward PE is only 22.2 times (currently 72 times)!
Combined with the benefit of yen depreciation, it will also help its performance realization. Currently, with a conservative estimate at a 1:150 USD exchange rate, the net profit for 2026/27 is very likely to reach 328 billion yen, corresponding to an EPS of about 180 yen. The forward PE will significantly drop to 22.2 times, which is very attractive compared to hardware companies with PE ratios often over 100 times!
By the way, my favorite Micron (#MU) has now fallen back, and its forward PE has dropped below 10 times. If it can be bought under $800, that would also be very attractive! 🧐



The three major technological revolutions of this era:
1. AI
2. Robotics
3. Quantum Computing
With AI being overheated in speculation, it’s worth studying the latter two!
Currently, the robotics sector hasn’t reached its GPT moment yet, and I believe its importance is on par with AI.
I often read @citrini’s memos, which can be said to be the most insightful research institution in the tech world, especially the dystopian scenario report "Autonomous AI Agent Out-of-Control Spread and the 2028 Global Intelligence Crisis," which directly dragged down SaaS software stocks at the beginning of the year—undeniable strength!
Today I saw Citrini invested in Robostrategy, which is an important signal. If you don’t know which robotics company to invest in, you might as well take a look at Robostrategy, stock code #BOT. It is an investment fund focused on the robotics sector and surged 11.8% against the market trend yesterday.
It has invested in many promising and unlisted robotics unicorn🦄 companies’ equity, such as Figure AI, Apptronik, Standard Bots, etc. Worth keeping an eye on🧐


Rocky_BTC
Today, Jensen Huang and female stock goddess @aleabitoreddit both simultaneously called out the robotics sector! Highly consistent, quite interesting! 🧐
The entire A-share robotics ETF surged 6% today, as shown in image 2. Now is the perfect time to position a bit for U.S. robotics companies tonight!
By the way, here is the complete humanoid robot supply chain map recently released by Goldman Sachs, given to AI to see its recommendations and corresponding stock codes: 👇
1️⃣ Humanoid Robot Manufacturers:
$TSLA (Tesla)
$XPEV (XPeng Motors)
BYD (002594)
Xiaomi Group (
UBTECH (
GAC Group (
Estun (
2️⃣ Foundation Models:
$NVDA (NVIDIA)
$GOOGL (Google)
$META (Meta)
$TSLA (Tesla)
iFLYTEK (
3️⃣ GPU / CPU:
$NVDA (NVIDIA)
$INTC (Intel)
4️⃣ Actuators and Dexterous Hands:
Harmonic Drive (6324.T)
Schaeffler (
Green Harmonic (
Sanhua Intelligent Controls (
Top Group (
Inovance Technology (
Beite Technology (
Hengli Hydraulic (
Shuanglin Shares (
Weichuang Electric (
5️⃣ Sensors:
$NOVT (Novanta - owns famous sensor brands)
$VPG (Vishay Precision Group)
Murata (6981.T - Murata Manufacturing)
TDK (6762.T)
6️⃣ Vision (Cameras / LiDAR):
$HSAI (Hesai Technology)
$INVZ (Innoviz)
$OUST (Ouster)
RoboSense ()
Orbbec ()
Sunny Optical Technology (
7️⃣ Batteries
CATL (
Samsung SDI (006400.KS)
Panasonic (6752.T)



