$TSM TSMC Tape Reports
Per Ticker.id: $TSM TSMC Tape Reports — 62 podcast mentions across 21 podcasts (30 days), latest 2026-08-26 16:00 UTC.
I wouldn't tie the two together. I don't think there's a pull forward. Compute is sold out and TSMC remains clearly the bottleneck. And so I don't see really that as an issue. I think, you know, the truth is, you know, NVIDIA has to kind of prioritize who they're going to deliver their compute to. And I think, you know, they take care of the hyperscale friends, but also they take care of their favorite neo-cloud friends. I think the recurring revenue stream is really more of, you know, the financing aspect. And in return, if those customers generate a greater return on those GPUs, they share or benefit, you know, in that greater return. And, you know, I think that's where, you know, you could start to see that in the dialogue. But, you know, the contracts that they've signed to date are de minimis. So I think it really would take time for that to be a meaningful driver for them.
CJ Muse — Squawk on the Street · 10AM Hour: Nvidia Ahead, Meta Settles, & PCE Comes In Hot 8/26/26 · 2026-08-26Yeah, so there's a bit of a— this is always a fun question, right? Which is where does the value go in AI? AI is generating all this value. You've got the end user, which I think we all agree is generating more value than anyone else, hence they're paying a lot for these models. But then you have, you know, the app layer. Well, so far the app layer has generated very little value. Then you've got the model layer, which again, up until, up until a year ago was generating negative gross margins and is now generating massive positive gross margins and looks like it's on the path to generating, you know, $100 million per megawatt. So turning, you know, $10, $15 into $100, as you said. But if we go back again a year ago, the hardware supply chain was generating all this gross margin while literally everyone else was losing money on it. OpenAI and Anthropic were just plowing VC money in, and as were many other startups. And many of these hyperscalers are building infrastructure without knowing if there was going to be a payoff. So ultimately you had this like, you know, negative value being created on the model layer almost, if you will, because they were selling the tokens for less than it cost them on the infra side and all the values being created, used at the chip, the fab. Initially in 2023, the memory guys were making no money. Off of, you know, HBM or memory for AI, even though theoretically their value they were delivering was humongous. Now you've got, well, actually TSMC makes way less value than the memory guys. Is that actually how much, you know, they're capturing less value, you know? So the value capture shifted around a lot, which is very fun for people tracking the market or participating in the market like Jane Street as an example.
Dylan Patel — Dwarkesh Podcast · Dylan Patel – Anthropic & OpenAI will have most of the world’s compute by 2028 · 2026-08-25And so I think if that's the case, right, then what happens to the price of compute? Well, if I'm Anthropic, incremental compute is worth it. Maybe I spend $40 million a megawatt on SpaceX compute. And if I'm SpaceX, you know, I look to the supply chain, I'm like, well, you know, I've struck this deal with Jensen where he's now all of a sudden using Twitter. And, you know, there's Elon saying they're exclusive to Nvidia, but Why doesn't Jensen raise his prices? And then, you know, SK Hynix and Micron and Samsung looked at Nvidia and were like, well, why don't they raise their price? So I think, I think the value capture, there's a bullwhip effect here, right? Where just because someone has risen the prices doesn't mean the entire supply chain rebalances immediately. Yeah, but over time the supply chain will rebalance and things will cost more and more. And, you know, to get that incremental capacity, you sort of have to, right? So TSMC raising prices very slowly, but memory companies raising prices very quickly. Substrate companies raising prices very quickly, different parts of supply chain raise, you know, Elon wouldn't have sold if it was $15, but he's selling because it's $25+. So obviously he rose his prices really quickly.
Dylan Patel — Dwarkesh Podcast · Dylan Patel – Anthropic & OpenAI will have most of the world’s compute by 2028 · 2026-08-25I do think in 2028 they have a big uplift in what compute they're able to deploy. 2026, they're still mostly relying on a lot of the smuggled chips, you know, you know, a lot of the chips that TSMC made. For companies that they thought weren't Huawei but ended up being Huawei, or a lot of HBM that Samsung is shipping, you know, sort of. But in '27, fabs start to go up. In '28 especially, fabs start to go up from SMIC and CXMT and such, where domestic production is actually reaching many millions of units a year. And now they're incrementally adding, you know, 5-10 gigawatts in just 2028 of domestically produced chips, right? Those chips are definitely worse than the chips that Nvidia will have in '28, or Google will have in '28, or OpenAI will have in 2028.
Dylan Patel — Dwarkesh Podcast · Dylan Patel – Anthropic & OpenAI will have most of the world’s compute by 2028 · 2026-08-25Yeah, well, first I wanna talk a little bit more about what's driving these gross margins. I mean, Matt hit on it briefly, but to understand, you know, how is Nvidia commanding these roughly 75% gross margins, you really have to look more at that supply-demand imbalance in high-end computing that we're seeing right now. So right now, the hyperscalers, right? Microsoft, Amazon, Alphabet, they're ordering these next-generation chips faster than Nvidia's manufacturing partner, TSMC. C can actually produce them. And, you know, it's a classic scenario of when demand heavily outstrips supply, you have essentially total pricing power. NVIDIA can pass these rising input costs, like the surging prices of high-bandwidth memory, uh, from suppliers like SK Hynix. They can pass these costs right onto their customers without hurting order volumes. But it's also important to note, you know, they have millions of developers locked into their proprietary, uh, CUDA software ecosystem. And, you know, building or optimizing an AI model for anything else takes months of engineering work. There's a real lack of viable alternatives that work out of the box. And so a lot of the tech giants choose to pay NVIDIA's premium prices rather than to risk falling behind in the AI race. And they're buying in so doing from NVIDIA really what's an entire ecosystem, not just the silicon. Now for me, where I would maybe start getting a little bit nervous, or at least questioning what's happening behind the scenes, is if gross margins were starting to fall down towards that 70% floor. It kind of might tell us a bit of a story about what's happening on the ground. It could indicate That supply would've caught up with or exceeded market demand. Could meet that— also mean that some of those cheaper competitive architectures, like AMD's MI300 series or hyperscalers' internal custom chips, which is another piece as well to consider, might have achieved some software compatibility that bypasses that moat. Now, I do not think that we are anywhere close to that reality. I also don't think, to be clear, that this is a winner-takes-all scenario. But those are some things to watch as we get deeper into the AI race and the AI revolution.
Rachel Warren — Motley Fool Hidden Gems Investing · 1 Earnings Report That Could Move the Market · 2026-08-24So CATL is dual listed. The A-shares trade in China with the ticker 300750. and the edge shares trade in Hong Kong with the ticker 3750. So same company, same one share, one vote, same dividend per share. But if you look at the stock price and after adjusting for the FX rate, you'll notice that the Hong Kong shares trade at a big premium of roughly 30 to 35% to the mainland China shares. And that's actually opposite to the norm. Normally for a dual-listed Chinese company, it's the mainland A-shares that trade at a premium because Chinese domestic investors face capital controls and they cannot freely buy in Hong Kong. So their demand gets concentrated into the local listing. But for CATL, it's the opposite case, and the reason is pure supply and demand. So the Hong Kong float is relatively small. And the demand from global investors is quite high. The Hong Kong and Chinese shares are not fungible, meaning that you can't buy in one market and freely convert and sell it in the other market. It's a bit like TSMC, whose US-listed shares have long traded at a premium to the domestic Taiwan shares. So if an international investor, if you have access to the Chinese shares, then that is a cheaper way to get access to this business. Some brokers provide access to domestic Chinese listings using something called the Northbound Stock Connect, but this is usually limited to the institutional investors. For smaller retail investors, Hong Kong shares are the only option. These are freely accessible to everyone, but you will have to pay a premium to own them. Now, CATL has been widening the Hong Kong float So after the Hong Kong IPO in 2025, they did a follow-on placement in 2026, and yet the global demand was so high that the premium hasn't compressed much. My personal view is that in the long run, Hong Kong shares should trade at a premium of about 10 to 20% range using TSMC as the reference. So TSMC has averaged at around 15% premium. But how and when that gap will compress is hard to say.
Mohnish Pabrai — The Investor's Podcast (We Study Billionaires) - The Investor’s Podcast Network · TIP840: CATL: Powering EVs, Power Grids, and AI w/ Stig Brodersen, Manish Karira & Ralph Summerford · 2026-08-23It strikes me that, you know, I think for people who are not deep in the weeds, they don't understand, I think, the margin stack that exists because the hyperscalers charge, I think, about 30%. I think their gross margin is about 30 to 40%. NVIDIA is up there at like 70%. The memory guys are at 80 to 90% now. And all of this stuff like stacks on top of each other, right? And when you look at how they end up stacking, and you know, you have at the very top OpenAI with, or Anthropic with a 70 to 80% margin, and they're buying tokens from Amazon with a 30% margin. Amazon's buying chips from, chips and other things from other people, and those people have like 50 to 60, 70% margins. Everyone then manufactures at TSMC and they have 50% margins and TSMC suppliers, ASML, they have 50% margins. And if you look at the margin stack, right? Like this $50 billion per gigawatt data center, it's really kind of made out of sand, literally, literally in some sense made out of sand. Sand and intellectual property. And when I think of it, it's all of that money is just the incentives required to get the humans, some of the smartest humans in the world to take a look at these problems and fix them, right? Like all of that money, like, because again, the physical elements inside that data center are actually worth not that much. Very little gold in there, very little gold and mostly silicon, some plastic. And if you just knocked down the entire data center and kind of sold it for scrap, it would be literally worth cents on the dollar, like few cents. And it just strikes me how much all of it is just intellectual property. It's really just know-how, intellectual property, It also somehow also strikes me how AI data centers are kind of this like crowning achievement of humanity as a whole in the sense that how many of these parts come from, you know, you have like argon gas from Ukraine and you have like copper from copper mines in Mongolia. And then, and all the way up the stack, you have like chips from China, rare earth metals from China, chips from Taiwan.
Speaker C — "The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis · AI in the AM — Weekly Highlights: Relaunch Week (Aug 17–20, 2026) · 2026-08-22Hey, I'm so excited to be here with you and Carl today to talk about this subject. And Carl and I had a great conversation about this at lunch the other day. So we're going to be hashing out some of those points here. And just to frame the debate at the highest level, 11 companies make up about 38% of the S&P 500. Actually, 9 companies make up about 38% of the S&P 500. In the mega-cap tech complex. These companies are Nvidia, Alphabet, Apple, Microsoft, Amazon, TSMC, which is not in the S&P 500 but is a major semiconductor manufacturing company, Broadcom, SpaceX, Meta, Tesla, and Oracle. And also call out that SpaceX is not yet in the S&P 500 following their recent IPO. But these 11 companies add up to about $32 trillion in market cap value, $33 trillion in market cap value, and $30 trillion of that is in the S&P 500, or about 40%. Of the S&P 500. And these companies are richly valued. I think the aggregate multiple is well north of 6 to 7 times of total sales for this aggregate. It gets worse if you exclude Apple, for example, as many people want to argue. And it gets even worse if you exclude the sellers like NVIDIA and TSMC. The question is, can these companies win? And what do you have to believe for them to win? I am skeptical of that, and I am so skeptical that I've made significant moves. I have paid taxes in order to do so, to shift my wealth away from the S&P 500 index fund to an equal weight index fund into factor tilts in small cap value, US international. Carl and Mindy have done the exact opposite and I think are bullish on the— this, this AI world and AI complex and think that there's many paths to winning here. And you have heavily invested your personal net worth in these stocks and continue to hold these positions as they've grown over the years. Is that the right way to frame it, Carl?
Scott Trench — BiggerPockets Money Podcast · The Bear Case vs The Bull Case for Megacap Tech · 2026-08-21