TSMC

TSMC is the ‘kingpin’ of chips and Apple has always been its best friend. That just changed

by

TSMC is the foundry of the world. Although there are others like Samsung that have muscleit is the Taiwanese company that has conquered the high-performance chip segment. It has achieved this through capacity, technology and an alliance: that of Apple. For a decade, TSMC was Apple’s great friend, the one that manufactured its chips and the one that revolutionized – with the designs of Apple Silicon– the laptops. Now NVIDIA rules. And he has elbowed his way through. In short. In the midst of the AI ​​era and with a technological current in which it is impossible to separate oneself from NVIDIA, Apple has more than enough reasons to feel jealous. While the mobile segment faces cuts unprecedented due to the crisis of RAM and components, and with Tim Cook himself -CEO of Apple- commenting on the difficulties they will have This 2026, artificial intelligence is going like a rocket. Major memory manufacturers have pivoted to high-bandwidth memory for AI GPUs, and companies like NVIDIA, Phison, amd and even the Chinese ones like SMIC and Huawei They are clapping their ears. They have made the AI ​​Big Techs dependent on their hardware, and no one makes that hardware like TSMC. Result? According to the latest reports, NVIDIA will become its largest customer this year. The importance of ‘Customer A’. It may seem like an unimportant change of chips, but it is actually more relevant than we think. The difference between a ‘Customer A’ and a ‘Customer B’ implies that, faced with production bottlenecks, one of the two is given priority. We already saw this in the 2020 semiconductor crisis when, precisely, half of the industry was drowning (cars, cameras, TVs and mobile phones) while Apple did not have such bad forecasts because it was the darling of a TSMC that was going to focus on iPhone chips to consolidate a lucrative relationship that began with the Apple A8 of the iPhone 6. Jensen Huang himself -CEO of NVIDIA- has commented the quite proud play on a podcast. “Morris –Morris Chang, founder of TSMC and friend of Huang – will be happy to know that NVIDIA is TSMC’s largest customer right now,” said the CEO. It is because little margin: 19% for NVIDIA compared to 17% for Apple, but it is an achievement and a thermometer of how the industry is doing. Last year, NVIDIA’s contribution to TSMC was 12%, which is a considerable jump in a very short time. “I need a lot of wafers”. Obviously, this does not imply that TSMC is going to stop pampering Apple over other companies. Apple has a huge percentage of the mobile segment, but NVIDIA is crucial to keep the AI ​​machinery rolling. Despite the Google attempts with its TPUsthe agreements of OpenAI with Broadcomthose of Goal with NVIDIA and AMD or those of xAI manufacturing its chipsNVIDIA is still the one who splits the cod. Even Chinese companies need NVIDIA GPUs and, of course, NVIDIA is more than willing to take a cut. On a recent visit to Taiwan,. Huang met with local industry heavyweights and noted that “NVIDIA would need a lot of wafers this year,” putting even more pressure to a TSMC that is crucial in the artificial intelligence chain. Synonym of success. Samsung, Huawei and SMIC They are fighting to be alternatives in case TSMC collapses. But TSMC has put us on the couch and has been looking at how to diversify the business for a few years. In Taiwan they maintain the heart and the muscle, but the plant in Europe – in Germany – is underway and they already have an operational foundry in the United States. In fact, there are plans to expand it because they have more and more clients who need a very specific product that works like a Swiss watch. But this has a B side: all the industry’s eggs are in the same basket. If TSMC fails, the house of cards can collapse. There is already some report that indicates that the American plant that manufactures for Apple, Intel, NVIDIA or AMD is overwhelmed due to a huge amount of orders. And there, precisely, lies the importance of being a client A… or a client B. Images | TSMC, NVIDIA In Xataka | SK is one of the chip whales and it is clear about one thing: not all the money in the world will satisfy AI’s hunger for RAM

SMIC is the Chinese TSMC and sums up all investment in AI

by

The semiconductor industry has pressed the panic button. Unlike the chip crisis of 2020the new one has not been caused by a ‘perfect storm’ and has a first and last name: artificial intelligence. The push for this technology and the rampant construction of data centers has caused a stock out of all computer components. And if the world’s leading companies have been caught on the wrong foot, the Chinese ones are no exception. To the point that SMIC has confirmed the lean times. The Chinese industry has activated “crisis mode.” Crisis (don’t bother…). At this point, introductions are unnecessary. The RAM market is broken because the main players (Samsung, Micron and SK Hynix) have focused on creating memory for data centers. The reports point to an NVIDIA that would ‘pass’ on launching consumer GPUs this year because you need all possible wafers to create GPUs focused on AI. And after the SSD price risethe following can be the processors. A few days ago, a report from Reuters noted that Intel and AMD were beginning to notify Chinese customers to sit tight for new shipments of server CPUs. Because we always talk about NVIDIA as if it were the only one that creates components for data centers, but the chips Threadripper from AMD and Xeon from Intel are key parts of servers. According to Reuters sources, AMD has warned of an increase in delivery times from eight to ten weeks. And if we look at an Intel whose sales in China represent 20% of its general income, they point out that the price of the fourth and fifth generation Xeons are being rationed and, in addition, increasing the price by 10%. As with other components, they are not enough. From hope to victim. From Reuters they point out that AMD hopes that its relationship with TSMC will allow it to maintain the supply chain. The problem is that TSMC has to be working hard to supply all its customers. NVIDIA is putting pressureQualcomm also needs its premium range mobile processors this year and they are the ones that make Apple’s processors. SMIC -Semiconductor Manufacturing International Corp- is the main Chinese semiconductor company. It is the one that in recent years has gained prominence defy US restrictions and provide high-performance chips to Huawei. It is also the great hope, together with Huawei itself, to develop a GPU that can cope with NVIDIA. Zhao Haijun is the co-CEO of SMIC and has come forward to be anything but reassuring. Double reservation. SMIC is winning more money thanks to demand, but production capacity is limited, to the point that he considers that the industry has entered “crisis mode.” “Customers are hesitant to place orders because no one knows how many memory chips will be available and, therefore, how many orders for phones, cars or other products they will be able to build,” stated to investors. In fact, analysis from Counterpoint reflect this crisis with a figure: 2.5%. This is what would have lowered the shipping forecast for Honor, OPPO and Vivo phones for the coming months. That leads manufacturers to book with multiple suppliers, something Zhao has compared with the “double booking” behavior that can be seen on an airline. “If one airline is fully booked, passengers will immediately book another, making total bookings appear inflated even though actual demand has not increased proportionately.” In essence, if the production capacity is limited, but the companies that need chips reserve several suppliers that draw from the same source, it seems that the demand is beastly, although, in reality, this is not so much the case. The result is the same: the system collapses. Car-free highways. The note of the double reservation is not the only pearl that Zhao has left. It is evident that, whether they can handle the demand or not, it is good for SMIC because Chinese companies eager for chips to develop their AI (of which They don’t stop releasing new models) are purchasing components for their data centersand the executive commented that “no one has really thought about what exactly those data centers will do, but companies would love to build the total capacity of the next 10 years in just one or two years.” “It’s like building train stations and high-speed highways even if there aren’t that many cars that need them yet” – Zhao Haijun Three billion dollars. But it’s about being there, having a place at the table where the conversation that, they hope, will shape the future is being defined. American Big Tech plans burn more than 650 billion dollars this 2026 alone, a brutal increase if we take into account that the investment was 400,000 million in 2025. But if we look at the global estimate, with Alibaba, Bytedance or Tencent in the equation, the estimate is three trillion dollars for the next five years. Crisistunity. In the face of crisis, opportunity. We have already commented that AMD, Intel or NVIDIA they are saturated. Also Samsung, SK Hynix and Micron on the memory boat, and in that scenario is where there are certain companies that can begin to carve out a space for themselves at an accelerated pace. SMIC is one of them, satisfying the demand of local customers, but memory manufacturers such as CXMT or YMTC that They had never painted anything in the conversationyou have a chance. We already know that PC manufacturers such as Asus, Dell or HP are considering buying CXMT memories and that Lenovo has already started to do so. And it is not only a window for Chinese companies: Intel has an adventure with the Japanese SoftBank to stand up to Samsung and its HBM4 memories. And even ByteDance would be working on your own AI chip. The problem is the same one we have been talking about for a long time: no matter how much they increase production and no matter how much new players appear, they are all playing in the same game, that of allocating most … Read more

Micron has emulated TSMC and is spending $1.8 billion on a RAM factory. Don’t clap yet

by

Taiwan is becoming one of the technological hotspots worldwide. If the country was already at the center of the technology sector because it is the home of TSMCwill now take on more prominence in the new era of AI. Your crown jewel is investing an astronomical sum in the United States and, now, the American Micron ends to close a $1.8 billion deal in Taiwan. And you can guess the goal. Keep feeding the data centers based on RAM memory. Micron. In recent weeks, Micron has been one of the big names in the technology sector. However, Crucial may sound more familiar to you. It is, or was, Micron’s brand for consumer RAM, but also for storage. Their products are very well regarded when it comes to assembling a PC in parts, but They turned off the tap at the end of last year and the last shipments will occur in February 2026. Now, Micron is shifting its focus to something much bigger and more lucrative: artificial intelligence. Specifically, supplying those same components, but to large companies that are setting up gigantic data centers. In the end, a data center It is made up of hundreds of “computers” that need both storage and RAM. The operation. Given the context, we come to the news. As the company itself has confirmedhave just signed an operation worth $1.8 billion to take over the P5 factory of the Powerchip Semiconductor Manufacturing Corporation -PSMC- company in Tongluo, Taiwan. An operation like this must pass several filters, but the company’s intention is for the transaction to be closed by the second quarter of this same 2026. They have stepped on the accelerator, and as soon as they can, they will begin to do one thing: increase the production of DRAM memory. clean room. Micron has confirmed that it is just one of the operations it is contemplating in a global expansion movement “to meet the long-term demand of its customers,” and acquiring a semiconductor factory makes perfect sense. Beyond the fact that the components and machines are different, there is something that factories of this type share: clean rooms. It is an extremely… well, clean facility stripped of any external elements. Suspended particles are kept at extraordinarily low levels, temperature, humidity and pressure are highly controlled parameters and the air is filtered numerous times per hour. Static electricity is reduced as much as possible and, ultimately, it is a clinical space so that no impurities interfere with something as sensitive as the manufacturing of semiconductors. It is, in short, like an operating room (or stricter if possible). Example of a clean room “All in one hour“Creating something like this requires a considerable investment (which is why new companies are entering to compete in the RAM segment, as rumored with Asusit is tremendously complicated), and that is why Micron has taken over existing facilities that they will only have to adapt to their activity. Besides, take the example of TSMC. In Taiwan, all components TSMC needs are “an hour” away. This allows the assembly line to be efficient, minimizing time, maximizing production and saving money. The new Micron factory will be very close to the one they already have in Taichung, being able to emulate that way of working that has led TSMC to excellence. Consumption RAM for when. Micron is expected to begin optimizing the manufacturing process in the new plant by the second half of 2027, but thanks to the context we gave before, we know that these “customers” are not those who want to assemble a PC in parts or even assemblers such as Asus, MSI, Lenovo or Gigabyte: they are the ‘Big Tech’ that are setting up data centers. In a recent interview, Christopher Moore, vice president of marketing for Micron’s client and mobile business, said the problem and the RAM bottleneck is elsewherebut also stated that this growth in data centers has gone from representing 30% of its market to 60%. He also stated that, although Crucial has disappeared, Micron will continue to supply memory to OEM manufacturers, but it is evident that the bottleneck is affecting, that prices are through the roof and that things are not looking good if you had to renew PC.E And, according to Micron’s vice president, it will continue until 2028. At least. Images | Maxence Pira, Hunter Trick In Xataka | Google doesn’t have rockets, but it is going to install data centers in space. SpaceX and Blue Origin rub their hands

Taiwan colonizing the United States with TSMC as the spearhead

by

TSMC is the big name in the global semiconductor industry. We all have companies like NVIDIA, Qualcomm, Intel or AMD in mind, but It is TSMC that produces most of the chips of these companies. The Taiwanese company produces around 60% of the world’s chips, but when we talk about the most advanced chips, that dominance is practically total. It is a technological candy that has decided to expand and, after the plant in the United States, continues to buy land to expand its footprint. And it is a move that further unbalances the balance in terms of chips. Necessary expansion. TSMC’s base of operations is in Taiwan, but a few years ago, the company saw clearly that they had to expand their operations framework. It is something that responds to a double need. On the one hand, the more footprint they have in other countries, the more the technology industry will continue to depend on their technique. On the other hand, the main factor: the threat of china. China and Taiwan are going through a period of growing tension. We have seen maneuvers by China that I’m sure they’ve made the Taiwanese nervous.. Also Taiwan operations for show that they could defend themselves and countries like Japan and, above all, the United States They are very aware of the situation. 87% of TSMC’s more than 80,000 employees operate in Taiwan and any open conflict between the countries would mean a stoppage in the company’s operations. Little joke with this: if its chips move the world, let TSMC stop producing would cause an economic collapse. Arizona. There is a third factor that is encouraging this international expansion. Although Europe, the United States and China seeks national sovereignty in semiconductor matterthe reality is that companies need the cutting-edge chips that only TSMC can reliably mass produce. And, while financing semiconductor plants, countries have decided to invest millions to attract TSMC to their territories. The plant that will open in Germany either that of Japan They are two examples, but the one that is already operating is the American one. Although Trump, with his protectionist policies and ‘America First’, does not like it being a foreign company that cuts the cod, TSMC already has a huge plant located in Arizona from which it produces key components of the iPhone 16. This facility is the company’s most ambitious project far from Taiwan, and what started as a $12 billion investment in 2020 has become a colossal $160 billion-plus operation. They started to produce 4 nanometer chips at the beginning of 2025 and the idea is refine machinery to reach 2nm in 2029. New lands. Within the ‘Made in the USA’ strategy of the large American technology companies, TSMC Arizona is vital. And considering the economic opportunity that the AI ​​era has opened up, with the astronomical need for chips to create products like NVIDIA’s solutions for data centers, TSMC wants to grab as much of the pie as possible. As we read in The Wall Street Journala series of factors such as Taiwanese investment and a relaxation in US tariffs on Taiwan would allow TSMC to expand further. According to the media, last week the technology company purchased 900 acres – about 360 hectares – of land adjacent to its current property in Arizona. The total has been almost 200 million dollars and the intention is to expand the facilities to reach a dozen. TSMC + NVIDIA Made in the USA (more expensive). With this move, TSMC would discourage the competition from trying to invest to stand up to them because, as we say, they are the ones who dominate the production of advanced chips and who have the capacity to supply their enormous customer base. Apple is one of those that already buys chips from Arizonabut NVIDIA has confirmed that its B30 GPUs will be the first made in the United States. Now, there is a toll. HE esteem TSMC Arizona prices on advanced nodes are between 5% and 30%. There are several factors. In Taiwan they have the policy of “everything at one hour”, so any material the factory needs is very close, creating an extremely efficient chain. That does not happen in the American factory, where suppliers are far away and you have to resort to air transportation, which increases the price. There is also the fact that the wages They are higher in the US than in Taiwan. Headache. Despite these conditions, and being a foreign company controlling the show on home soil, TSMC has so dominated the process that the companies it’s worth it because they know that the chips they get will be the best for their products. Furthermore, from a political perspective, these additional costs may even be reasonable if they ensure that a conflict in Taiwan would not completely paralyze its economy. For TSMC, expansion is a great move. At the political level, countries that embrace their factories also have a reason to attract investment and Big Tech and the CEO of NVIDIA himself is clear that swith those who will lead the industry for decades. However, it is still an industry dependent on a single entity. Without leaving the United States, the country got his hands on Intel in the middle of last year in an almost unprecedented move to turn the company into the great american foundry. With TSMC expanding its network at home, they are going to have it complicated despite having the best technology available. Images | NVIDIA, TSMC, Intel In Xataka | The world’s technology industry practically depends on a single road: the one that leads to the Spruce Prine mine

The US has insisted that TSMC manufacture chips in Arizona. The reality: it is a disastrous idea

by

TSMC, the world’s largest semiconductor maker, has long been pushing for unprecedented expansion outside Taiwan. The initiative includes large projects in the United States, Japan and Germany, but does not respond to market demand, but rather to geopolitical pressure and a chip war that wants to try to “repatriate” this type of process. It’s a terrible idea. Morris Chang knows it’s a mistake. Despite the political urgency, the economic viability of these factories abroad has been questioned by TSMC founder Dr. Morris Chang. He already had the previous experience with the WafertTech factory in the US in 1996, and has qualified Arizona initiative as “a very expensive exercise in futility” Everything one hour away. Chang’s skepticism is based on the belief that TSMC’s operations and profitability are intrinsically dependent on its ecosystem, which is entirely concentrated in Taiwan. The Hsinchu Science Park “cluster” allows hundreds of technology partners to operate within a “one-hour” radius, facilitating problem resolution and providing ultra-fast logistics and unparalleled coordination. TSMC is still 90% Taiwanese. Despite that global expansion, TSMC remains deeply Taiwanese, with more than 90% of its manufacturing capacity and nearly 90% of its employees on the island. That’s where your massive, highly trained and qualified engineering talent base is. That is again a key factor in its competitive advantage, and in fact the company has already warned its employees in the US that they should adhere to the work culture of the Taiwanese company. Arizona produces, but it is more expensive. That attempt to replicate Taiwanese efficiency in Arizona has revealed something important: although TSMC has achieved competitive performance in its first production runs with 4nm photolithography, the cost of the wafers is significantly higher. The local supply of raw materials and equipment remains insufficient, making the factory dependent on Asia and is a bottleneck for the efficiency of the production cycle. Skilled labor shortages and permitting and bureaucracy, which further slow things down, add considerable operational costs. Japan and Germany, next objectives. TSMC has two major expansion projects in Japan (JASM) and Germany (ESMC). These locations will focus on much less advanced photolithographic nodes (28/16 nm) and will focus on meeting the demand of some specialized customers such as Sony for image sensors in Japan or Bosch in Europe. The scale of these investments is less than that of Arizona, which aims to be the world’s largest advanced chip factory… if planned future phases are completed. A double edged sword. TSMC’s expansion has two sides. On the one hand, TSMC consolidates its technological leadership and its strategic role as a “silicon shield” against China. On the other hand, it generates internal anxiety about the possible “leakage” of advanced technology and talent that could weaken national sovereignty in the long term. US pressure even extended to veto the possibility of establishing a TSMC factory in the United Arab Emirates. TSMC does not expand by pleasure, but by pressure. Traditionally, TSMC only builds new factories in response to real demand from its customers. Here the reason has been very different, and geopolitical pressure has forced moves that the company would probably never have made otherwise. Here the different subsidy programs (CHIPS Act in the US, European Chip Law) try to repatriate part of the manufacturing and thus mitigate Asian dependence, but it’s not clear at all that they achieve it. Image | TSMC In Xataka | Japan is rapidly reconquering the chip industry. It has just successfully manufactured its first 2nm transistor

Huawei lost to Google, Qualcomm and TSMC. What he didn’t lose was something more important: his reputation.

by

Last week were the Xataka Awards 2025. Stella Li, global vice president of BYD, took the Xataka Legend. He Galaxy S25 Ultra It swept the super high range. Freepik was crowned as best Spanish technology company. It was a night of proper names, drinks and conversations with readers. But There is a prize that, for those of us who spend a lot of time in Xatakaas workers or as readers, has a special weight. Not because of its glamour, but because of what it represents. The Community Award is not decided by any jury. There are no internal debates. You, the readers, decide with your votes. It is the device that you liked the most, without filters. In fact, it is the only one that is not delivered by any employee of the house, but rather by members of the community who represent it on stage. In the image that heads this article, three of them with Cristina Isidoro, PR Manager of Huawei in Spain, who collected the award. Because this year he won it Huawei Watch GT 6 Pro. And when I saw the result, I smiled slightly: it was more than just a reward for a well-made smartwatch. It was pure symbolism. Look at the historical list of winners of this award: Almost all, Chinese devices or devices with a Chinese soul that share a pattern: focus on value for money, practical innovation, and in some cases, arriving wanting to break molds. But among all of them, Huawei is the only one that did not arrive yesterday promising a lot for little. It is the only one that was already in the world elite, disputing the throne with Samsung and in fact about to snatch it awaybefore the United States decided to use it as a pawn in its trade war. Because Huawei has not conquered the perception of premium quality by offering more gigabytes for fewer euros. It conquered it by being, for years, simply a great option. He P20 Pro It was the first mobile phone with a triple camera that really worked. The Mate 20 Pro was an unapologetic technical beast compared to the high-end greats. Their MateBook laptops have been worthy rivals of the Surface. And their GT watches already stood out for batteries that lasted weeks when Apple asked for a charger every night. They weren’t cheap. They were good. And that difference, in the technology market, is abysmal. Then 2019 arrived. EntityList. American veto. Goodbye Google, goodbye Qualcomm, goodbye TSMC. Sales outside China plummeted and the Western narrative was unanimous: Huawei was dead. Without the Google ecosystem, without access to the supply chain, it was impossible to survive in this business. But no one explained to them that it was impossible. Instead of giving up, they built their own universe. HarmonyOS on more than a billion devices. Kirin Chips own, then Ascend for AI. Huawei Cloud growing in Asia, Africa and Latin America. They didn’t beg to go back to Google Play like we might perhaps have expected them to do. They simply built another entire ecosystem. Without one word higher than another. At the beginning of the month I was in China and was able to try several of their devices, including some that just left there. The premium feel is real. And something that we do have here, the GT 6 Pro, is not a gadget 150 euros that promises too much and falls halfway. It is a watch in the 400 euro range that performs very well. and the community of Xataka has passed sentence with his prize. That doesn’t happen by chance. Xiaomi shines in value for money. Realme and Oppo play there. Nothing has its aesthetic indie. But Huawei is the only Chinese brand that, when you mention it, the European consumer automatically thinks of “serious quality”, without the asterisks that others have. And she did it right after they tried to destroy her. The Huawei Watch GT 6 Pro is a great watch. But winning the 2025 Community Award means something else: It is the recognition of the only Chinese brand that has come out of the perception low cost without giving up its origin. It is the prize that, in a way, China had been pursuing for decades. Respect without conditions. And it has been won by a company that they tried to annihilate. Sometimes vetoes don’t kill. They forge legacies. Featured image | Xataka In Xataka | The LG OLED Signature AI T4 is the best television of the year for a simple reason: we are saying goodbye to the black monolith

ASML and TSMC are masters of semiconductors. A US startup believes it knows how to end them: with X-rays

by

A mysterious American startup called Substrate has made its appearance with a purpose extraordinarily ambitious: compete head to head with ASML. The Dutch company has become the master of the segment of advanced photolithography machines for chip manufacturing, but at Substrate they believe they have the key to turn the tables. Why is it important. ASML has no competition in the market since it placed its first equipment of UVE photolithography. The ASML thing is a monopoly de facto: If a chip manufacturer or designer wants to access to produce the most advanced models, it depends entirely on the Dutch company. No one has managed to stop it since then, and even China, which is trying to free itself from that dependence, it’s really complicated. Substrate. This is the name of this startup that has developed a new team in which use particle acceleration to manage lithography. This technology allows microscopic circuits to be etched onto silicon wafers, and this new company claims that its machines could be in manufacturing plants in the US within the next two years. It all sounds very good. Maybe too much, but they already have funds to try: they have just raised 100 million dollars and among investors There is Peter Thielco-founder of PayPal and current CEO of Palantir. And already, they will try to help create the new TSMC. The challenge is enormous: ASML has invested decades and billions of dollars to perfect its photolithographic equipment, and the complexity of this market makes it very difficult for companies created from scratch to compete. Substrate’s objective is twofold, because it also aims to ensure that its machines enable the affordable chip manufacturing in the US. Or what is the same: it not only wants to compete directly with ASML, but also allow American manufacturers to compete with TSMC on American soil. There it is nothing. Light is everything. When creating those circuits, some of the lines created are so fine that their dimensions are even thinner than the wavelength of light. To solve that ASML problem they do use of extreme ultraviolet light (EUV) through a very complex set of lenses and mirrors. These machines generate a type of artificial light that is capable of engraving those very fine lines on advanced chips. Very special X-rays. In Substrate they propose a different idea. Although they do not give all the details to prevent someone from copying them, their machines use X-ray lithography (LRX). A particle accelerator creates a light source from x-rays with a shorter wavelength, allowing the beam to be used to create advanced chips. Current Substrate machines are currently capable of working with 12 nanometer nodes, comparable to ASML High NA EUV machines: That would put them on par with the most advanced chip production lines in the world. LRX technology is not new. This technique has existed since the 1970s, but had been abandoned because longer wave techniques (UVP and UVE) continued to scale without the need to overcome the great technical challenges of RXL. In substrate they seem to have precisely the solution to those problems, which focused on optics and the need to use massive particle accelerators as a light source. And they promise brutal cost savings. At Substrate they maintain that their LRX equipment will have an estimated cost of about 40 million dollars, compared to 400 million for ASML’s High-NA EUV. In addition, there would be another associated revolution: the possibility of carrying out the so-called single exposure patterning at advanced resolutions (2 nm, 1 nm and beyond), thus eliminating the additional costs of the multipatterning (multiple exposures). If that promise is fulfilled, the production cost of advanced wafers would be 50% lower according to Substrate. But. Of course, it’s one thing to say it and another to do it. The precision of the process, they explain in Semianalysisis a monumental obstacle. The transfer of patterns (going “from light to silicon” almost “carving it”) remains problematic, and among the challenges is solving the roughness of the edges of the printed patterns, which is amplified during engraving. There is also the problem of stochastic noise (shot noise, random fluctuations in photons that cause defects) and secondary electron blurring inherent to the high energy of X-rays, which fundamentally limits resolution. There are also currently problems with the 1.6 nm overlay, which is still high for more advanced processes where 1.0 or 1.2 nm are sought. In Xataka | AMD is today a semiconductor giant. In its beginnings it was something much more humble: a blatant copy of Intel

OpenAI, Broadcom and TSMC want to end their XPUS

by

Openai wants to hit the table in the race to dominate artificial intelligence. At the beginning of September we knew that “a mysterious client” made an investment in Broadcom of 10,000 million dollars for the design of AI chips. Through sources close to the company, we learned that it was OpenAI, with the intention of form a strategic alliance with both Broadcom and TSMC for the manufacture of own chips. Called ‘Xpus’, they have a clear objective: stop depending on Nvidia, something that Many great technology want. Have independence in AI is paid expensive. For Openai, depending completely on Nvidia to feed chatgpt and its future models is a capital risk, especially for having to deal with astronomical costs, and being limited to the technological evolution and stock of a company like NVIDIA. Creating your own chips allows OpenAI to optimize hardware specifically for your language models, reduce operating expenses and free yourself from the bonds of a single supplier. A strategy on the Apple line when He said goodbye to Intel to start making their own chips in their Mac. Although stop depending on Nvidia today is complicated. The golden trio behind the XPUS. Broadcom Lead the design Of these specialized processors, contributing their experience in custom chips for technological giants. The company controls approximately 70% of the Custom IA processors market and has already collaborated with Google in His tpus For years. TSMC, the world’s largest semiconductor manufacturer, will be responsible for producing chips using their advanced 3 nanometers technology. Mass production would predictably start in 2026with the first shipments arriving that same year. What are exactly the Xpus. These chips are not adapted gpus for AI, which is what is broadly marketing nvidia, but architectures designed from scratch With a single purpose: accelerate language models. While the NVIDIA GPUSs were born to rendering graphics and then adapted to artificial intelligence, OpenAi XPUS are specifically conceived for training and inference tasks of AI. They incorporate systemic matrices, high bandwidth (HBM) and integrated network capabilities, similar to the most advanced NVIDIA processors but optimized for the specific needs of OpenAI. The rebellion against Nvidia has several fronts. Openai is not alone in his efforts to break the hegemony of Nvidia. Google has been developing its TPUS (Tensioner Processing Units), which already go for its seventh generation. Amazon created its chips Trainium and Inferentia For AWS. Microsoft designed Maia For Azure. Goal collaborates with AMD In alternative solutions. Even from China, Huawei threatens face face In the next three years. It is clear that all large technological ones want to have greater control of their products, and for that they must go down to the chip supply chain to start rethinking the scheme. Nvidia’s real pit. Despite how the panorama is, Nvidia preserves An ace in the manga That will not be easy to overcome: CUDA. This development platform has become the industry standard for more than a decade. Virtually all the researchers and developers of the CUDA program, which generates a very powerful network effect. Changing to a different architecture not only implies acquiring new chips, but rewriting software, removing equipment and, in many cases, starting from scratch. This technology is the one that keeps Nvidia in a privileged position even before the avalanche of competitors. A dual strategy. The paradoxical of the case is that Openai is not abandoning Nvidia. Parallel to this own chips project, the company maintains An agreement of 100,000 million dollars With the Santa Clara giant for its Stargate project, which plans to build huge data centers until 2028. On the one hand, the calculation power supply with NVIDIA is ensured, while on the other they develop their long -term alternative. The idea is that XPUS are initially used internally, especially for inference (apply already trained models), while NVIDIA GPUS will continue to be necessary for training the most demanding models. The board is removed. Broadcom is clearly reinforced from this movement. The announcement of your agreement with Openai He shot his actions In the stock market and consolidate its position as a preferential partner for companies that seek custom chips. TSMC also wins: each new wave of specialized chips reinforces its indispensable role as a world reference manufacturer. Nvidia, on the other hand, saw his price slightly fall after knowing the news, although he continues to maintain A dominant position Very difficult to collapse. In addition, AMD could benefit if smaller companies, unable to develop their own chips, look for more accessible alternatives than NVIDIA. Cover image | NVIDIA, VECTEEZY, Village Global In Xataka | The companies of AI tell us that they want to achieve an AGI. What they are really conquering is the economy of attention

The undisputed winner of the aggressive competition of TSMC, Intel and Samsung is a European company: ASML

by

Integrated 2 nm circuits are about to disembark in the market. Users know that nanometers have lost much of their usefulness, and that, in reality, They represent a category of semiconductors. In fact, they no longer faithfully reflect the length of logical doors or other physical parameter, such as the distance between transistors. Each chips manufacturer He manages them with freedomwhich prevents us from directly comparing the lithographs that try to “sell us.” Whatever the important thing is that TSMC, Intel and Samsung are about to engage in a new battle that seeks to capture the maximum possible number of customers for their 2 Nm or comparable line nodes. Whatever happens we can be sure that the great beneficiary of this contest will be The Dutch Company ASML. And it will be because it is the only manufacturer on the planet that produces the equipment of extreme ultraviolet photolithography (UVE) and haute opening that are necessary to go beyond the 2 Nm reaching the optimal performance. Digitimes Asia He has just confirmed that those responsible for the Samsung semiconductors manufacture are weighing the possibility of increasing the number of Uve haute opening machines that will buy at ASML. And, according to this Asian medium, it will do it because it needs to reduce the technological and commercial gap that separates it from TSMC, which leads the chip market with A fee close to 60%. The Uve High Opening machines are still in the test phase, but there are no doubt that they will be the authentic protagonists of the semiconductor industry in 2026 and successive years. ASML Haute Opening Lithography Machography is an engineering prodigy It weighs as much as two Airbus A320 and incorporates more than 100,000 pieces, 3,000 cables, 40,000 bolts, and also more than 2 km of electrical connections. The photolithography team Twinscan Exe: 5000 Designed and manufactured by ASML is the most sophisticated integrated circuit production machine that exists. And also the most expensive. The most up -to -date information we have reflects that only one of these teams costs 350 million euroswhich will surely cause some chips manufacturers think twice Before buying it. ASML plans to deliver to its customers annually from 2025 about 20 Uve Haute Opening teams ASML engineers have invested a decade in the development of the technology necessary to set up this machine, which, in reality, is a team of extreme ultraviolet lithography (UVE) second generation. This company of the Netherlands plans to deliver to its customers annually From 2025 about 20 teams of this type with a purpose: put in their hands the possibility of producing chips of 2 nm and beyond. Interestingly, to develop this machine, ASML engineers have made a very advanced optical architecture that has an opening of 0.55 compared to the 0.33 value that the first -generation UVE lithography equipment has. This refinement of the optics allows to transfer to the wafer patterns of greater resolution, hence it is possible to manufacture chips using more advanced integration technologies than those currently used in the nodes of 3 Nm. However, this is not all. ASML has also improved the mechanical systems that are responsible for the manipulation of wafers with the purpose of making it possible for a single UVE Machine to be able to produce more than 200 wafers per hour. The cover photography of this article allows us to intuit the extreme complexity and sophistication that one of these teams has, which, by the way, would not be possible without the cooperation of other companies, such as the German Zeiss or Cymer, a company of American origin that is currently consolidated within the ASML structure. Somehow this last company Delivery to ASML the raw material that need their photolithography machines. And that raw material is none other than the ultraviolet light that is responsible for transporting the geometric pattern described by the mask so that it can be transferred with great precision to the surface of the Silicon wafer. Image | ASML More information | Digitimes Asia In Xataka | The great covered in the War of Critical Minerals is Tungsten. The US needs it and 83% have it China

China’s last US hint threatens a TSMC chip factory ahead

by

On December 31, it will be a very important day for semiconductor manufacturers that have plants in China. From that date they will not be able Its facilities in this Asian country. And they cannot do it because The US does not want chips manufacturing equipment that resort to American technologies and innovations They arrive in China. Not even integrated circuit factories that do not belong to Chinese companies. In 2022 the US Department of Commerce granted a temporary exemption to several manufacturers of foreign semicondators who have plants in China so that they could equip their facilities with the machines they needed. But this permissive period is about to expire. From now on any chips manufacturer who has plants in China will have to request a license from the US Commerce Department to be able to install in its factories machines with US components or technologies. Intel has sold Your Dalian plant (China), so this measure no longer affects it. However, there are three foreign companies of enormous relevance in the semiconductor industry that will be affected by this measure of the US government: South Korean Samsung and Sk Hynixand the TSMC Taiwanese. The latter has a chips factory in Nankín, in the province of Jiangsu (China), in which as of December 31 it will not be able to install advanced lithography equipment. The US and TSMC strip and loosen The semiconductor production plant that TSMC has in Nankín is important for this company, but it is not a toe. In fact, it manufactures mostly chips in its 16 and 28 nm nodes. This installation currently represents only 3% of TSMC’s total production capacity, but this does not mean that it is not relevant within the manufacturing infrastructure of this Taiwanese company. In fact, in 2021 announced an investment plan of 2,870 million dollars that in 2023 allowed expanding the manufacturing capacity of the plant to about 40,000 wafers per month. These presumably “restrictions” will condemn “in the short and medium term to this factory to the production only of mature chips During the last weeks, the TSMC Directive dome has met with the US Department of Commerce in an attempt to protect the interests of its Nankín plant, But it has not been successful. These presumably “restrictions” will condemn “in the short and medium term to this factory to Production only with ripe chipsalready long term will probably lose its relevance in the Integrated Circuite Production Infrastructure of TSMC. Whatever this is only One more episode in the awkward relationship that support the US and TSMC government for years. For this chips manufacturer the country led by Donald Trump is very important because a good part of his best clients is American. Nvidia, Apple, AMD, Broadcom or Qualcomm, among other companies, get the chips they design in TSMC’s lithographic nodes. However, this currency has a second face. And it is currently the USA cannot do without TSMC. Intel is American, and It has advanced lithography nodesbut the competitiveness of his Taiwanese rival is difficult to match. TSMC has cemented its leadership on the tuning of a range of Very advanced high integration technologiesand, at the same time, On a colossal production capacity which is only possible reaching a very high wafer performance. The US government knows very well the strength of this company. And also how important it is for US companies that I have mentioned in the previous paragraph. Image | TSMC More information | SCMP In Xataka | Intel was about to snatch Apple as a client from TSMC. Having achieved its story would be another

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.