只分享个人观点。
LipBu Tan$Intel(INTC.US) You don't need to read this kind of news. The so-called process technology 14A is just a code name used by different companies. Samsung has long been mass-producing GAAGET 3nm, which is actually significantly inferior to TSMC's n3 node. Samsung is also the most notorious for inflating its process node claims.
A simple fact:
Process Technology: TSMC > Intel > Samsung
Foundry Experience: TSMC > Samsung > Intel
Intel
USINTC
$Intel(INTC.US) Please stop selling, I need to support my family
$Intel(INTC.US) Why is Simplified Chinese an information entropy lowland, with garbage information flooding
Intel
USINTC
$Intel(INTC.US) Don't be afraid, the semiconductor index is not doing well. I've always held a view: Intel's risk isn't in itself; the day cloud providers' capex can't keep burning, the entire upstream semiconductor industry will fall like sand and mud. However, OpenAI/Claude aren't listed yet. The day the primary market needs to cash out, that's the end. Personal opinion.
Intel
USINTC
Real estate can at least be lived in. Now, AI infrastructure lacks GPUs, domestic hardware, and hardware updates quickly, soon turning into massive waste, or it's already outdated when it comes out.
$Intel(INTC.US) 107 is around the support level, it was dragged down by the broader market last Friday.
Intel
USINTC
$Intel(INTC.US) is not considered a product for Apple's foundry business. Crescent Island and Diamond Rapids are likely the two most important products to be launched in 2027, addressing the current demand for inference and agentic capabilities. If Intel can capture a share of the inference computing market, its current valuation is still relatively low. Of course, there is significant uncertainty. Qualcomm previously developed an AI100 inference chip, which was a complete failure.
Intel
USINTC
$Intel(INTC.US) At 1:30 PM, can Lipbu Tan save it? 😭
Intel
USINTC
$Intel(INTC.US) Isn't that certain patriotic company recently bragging about some law of vertically stacked transistors? Take a look at the now-listed Clearwater Forest. Dozens of chips are stacked vertically/horizontally. A single CPU die is only 50+ mm², much smaller than a mobile phone SoC, with L3 cache of different processes stacked at the bottom. Chips are assembled like building blocks. It can be said to be the most advanced product in terms of packaging and process that is currently listed in the semiconductor field. (TSMC's N2 will only be available in the second half of the year, and CoWoS is horizontal stacking.)
Intel
USINTC
$Intel(INTC.US) advanced packaging, data center, foundry services. I've been talking about these three catalysts since last year. Now they've all been fully realized. Will start selling shares next month.
Intel
USINTC
$Intel(INTC.US)Intel is massively purchasing chip packaging equipment; EMIB must have gotten an order, I guess it's from Google. Finally, to show off a bit, over the past year, several of my predictions have basically come true. 😎
Intel
USINTC
$Intel(INTC.US) I've been waiting for news about Apple's A-series chip foundry business since last year. The reason is simple, let's think from common sense.
Known:
1. Ming-Chi Kuo previously had a bearish view on INTC, indicating his neutral stance, and considering his past leaks, his report that Apple is considering using 18AP for the low-end M-series models is extremely credible.
2. Using TSMC/INTC dual sourcing would increase Apple's chip backend design costs, and there doesn't seem to be a clear benefit.
3. Apple once used Samsung/TSMC dual sourcing at the 14nm node.
Based on the above, the answer is obvious. For a company of Apple's caliber, it would rather increase design costs and also needs to introduce multiple suppliers to reduce its own supply chain risks. The 18AP version of the M-series foundry business is just a trial run for both parties, equivalent to a test order from Apple to support INTC as a second supplier. Apple's chip design capabilities are also top-tier globally. If the 18AP process performs well, the real goal for both parties should be: the iPhone A-series on the 14A process. The A-series is Apple's most core chip product. If INTC can secure it, it's a foundational step on the path to a trillion-dollar market cap.
$Intel(INTC.US) I haven't added to or reduced my position, holding from over $20 until now😃. A normal person would have run away long ago. Last year, I set three core narratives: data center (now largely priced in), packaging (waiting for Amazon/Google news), and foundry (waiting for news on an iPhone A-series chip).
$Intel(INTC.US) If Apple wants to mass-produce the M7 series in H2 2027, the process selection is most likely already finalized now, but it can't be disclosed due to trade secrets. The package selection can wait until the IC front-end design is completed, so the Google EMIB selection is most likely in H2 this year, and they might have already received it now. Tesla's Musk himself has already said it, but the specific node is unknown.
These are tickets to a trillion-dollar market cap, but my 'Qilin Arm' is about to act up, I can hardly hold on anymore.
Intel
USINTC
$Intel(INTC.US) The market is just too ridiculous, either overly pessimistic (INTC with an $80 billion market cap) or overly optimistic, especially since it's still in the red on the books.
$Intel(INTC.US) had a gamma squeeze, burning the shorts😅
Intel
USINTC
$Intel(INTC.US) has secured an order from Musk for its 14A, making a market cap of 500 billion very likely. A few more advanced packaging customers could be the catalyst. Maybe Apple and MediaTek are also in the mix for the 14 series.
$Intel(INTC.US) It's not too late to buy now, though you might be trapped in the short term. In the next three to five years, the market cap could range from 400-500 billion to one trillion.
