chip

Nvidia has just presented the definitive chip against Intel and AMD. There is a problem: Windows

by

The Nvidia processor for PC is the “the wolf is coming” of consumer technology. The company has been the reference for years in GPUs for gamers and flirted with SoCs thanks to the Tegra chips (which are what give life to both nintendo switch like to nintendo switch 2), but for computers they still couldn’t find a way to get equipment with 100% Nvidia guts. That just changed with the presentation of RTX Spark chips. It is a SoC that directly attacks the binomial Windows PC = Intel or AMD CPUone that is positioned as the alternative to those traditional options and that is specifically designed to compete for the heart of the consumer PC. Specifically, for laptops. Now, although Microsoft and Nvidia have been generating excitement for a few days and pointing out that it is the new era of the PC, there is a problem. Windows. The brake is no longer silicon, it could be Windows The theory is very interesting. RTX Spark combines a CPU Grace up to 20 cores that it has developed together with MediaTek (this is curious) with an RTX Blackwell GPU with 6,144 cores. TSMC (how could it not be otherwise) has given life to chip in a 3 nanometer lithography. Not only is it powerful, but it has up to 128 GB of unified memory (the same design that we see in Apple Silicon) and an interface NVLink which allows communication between RAM, CPU and GPU to be very, very fast. Nvidia talks about rendering heavy 3D scenes on laptops, running models with 120 billion parameters, and at the same time running games at 1,440p above 100 FPS with DLSS and ray tracing. The best? That Jensen Huang stood out at the Computex conference showing this in very thin and light laptops. It is the same strategy that Qualcomm follows. own Microsoft has already presented a Surface with RTX Spark and it is an architecture that makes a lot of sense in the universe of current light but powerful laptops… and also in desktop computers like a mac mini or of a mac studio. And, compared to the more traditional PC industry, the GPU is estimated to be in the range of a RTX 5070 for laptops. In the absence of testing it, it is undeniable that it looks good and that, although there are data that are not so favorable (such as bandwidth when compared to the most powerful Apple), it is a good addition to a segment in which, if we left the Intel/AMD duo, the only one that was trying was Qualcomm with devices like the Snapdragon X Elite. And there is the key: RTX Spark, like Qualcomm chips, is focused on being the heart of a Windows that is at its brightest. Because RTX Spark is a chip with ARM architecture and, although in office tasks Windows ARM It moves well, under more demanding tasks is when it begins to not be up to par. Microsoft’s system, which they themselves know is not at its best level of popularity due to the whole issue of AI features, has many shortcomings in its ARM version when it comes to gaming, precisely what Nvidia is promoting. It is also not the best optimized on laptop computers, something that is being seen with type machines. Steam Deck. The heart of the new Surface We are seeing it in recent years with PC-console asus, MSI either Lenovo: The hardware is good, but Windows drags down the experience significantly. The paradox is that the Steam Deck, being the least capable on paper, is usually more recommended precisely because it avoids Windows and relies on a system much more fine-tuned for that format. With RTX Spark, the two companies say they have been working for a long time to solve those problems and make this time, Windows on an ARM chip feel different with support for games with anticheat and native for personal agents. We will see in practice what ends up arriving, but two things are clear here. The first is that Microsoft gains aggressive hardware to compete face to face against Apple in the field of very powerful laptops with long battery life. The second is that Qualcomm is no longer alone in that corral and now it will be very interesting to see what hardware it responds with. Because Nvidia already has the chip, the CUDA ecosystem and agreements with all manufacturers, as well as the support of the giant TSMC. The “weak” link, therefore, is not silicon, it is a Windows on ARM that has improved a lot in recent yearsbut that is the element that will have the most to prove. In Xataka | Graphic muscle for Windows and a slam of the door on Android: the exclusivity toll that Nvidia demands with its new ARM architecture

The US opened the door to Nvidia’s H200 chip in China. The Chinese army has been waiting for a long time on the other side

by

Jensen Huang, the CEO of Nvidia, has been forced to “fight” with the US Department of Commerce for months, but he has achieved what he wanted: your company can now deliver some of its Chinese clients its chip to artificial intelligence (AI) H200. As we explain to you On May 14, Alibaba, Tencent, ByteDance and JD.com are four of the ten Chinese companies that already have access to this powerful GPU. And they have it because the US Department of Commerce, which is the institution that grants or denies export licenses, has authorized at least ten Chinese companies and several distributors, including Lenovo and Foxconn, to acquire the second most powerful AI chip that Nvidia has. This decision has come almost two months after the US Government confirmed which was going to allow the company led by Jensen Huang to deliver its H200 chip to its Chinese customers. However, Nvidia likely won’t have time to savor this victory. Once again, dark clouds are gathering over it that threaten to compromise, once again, its business in China. And, according to Bloombergat least seven Chinese universities linked to the country’s armed forces and defense industry are trying to obtain H200 chips. This disclosure comes from China’s public procurement records, so it is presumably reliable. Remote rental: the avenue that the Department of Commerce still does not know how to close In the US there is a pressure group that opposes the sale of advanced American AI chips in China. Chris McGuire, senior fellow on China and emerging technologies at the Council on Foreign Relations, holds that “any deal that allows Nvidia to sell more chips to China means fewer Nvidia chips for US companies and a smaller US advantage over China in AI.” Besides, McGuire argues that “it is surprising that President Trump continues to allow himself to be convinced to put Nvidia’s interests before those of America.” Chinese entities increasingly resort to renting airtime on servers equipped with restricted Nvidia chips What is happening right now with Chinese universities is the ideal breeding ground to reinforce the theses of this pressure group in the US. Two of the institutions that have expressed interest in H200 chipsBeihang University and Northwest Polytechnic University, are among China’s “Seven Sons of National Defense”, a select group of universities dedicated to supporting the People’s Liberation Army. Both have been included in the blacklist of the US Department of Commerce for their involvement in the advancement of Chinese military capabilities. And public procurement records reveal that the Beihang School of Cyber ​​Science and Technology, which claims to have “national defense characteristics and aerospace advantages,” is attempting to rent the use of Nvidia chips. Northwestern Polytechnic University’s School of Cyberspace Security is also trying to rent access to H200 chips, according to those same records. Chinese entities are increasingly resorting to time of use rental on servers equipped with restricted Nvidia chips as a way to access prohibited hardware without having to import it directly. This is the strategy that the US Government will surely try to dismantle. What is not clear at the moment is how he is going to do it. Image | Nvidia More information | Bloomberg In Xataka | The US remains committed to stopping China. Now it has targeted the second largest Chinese chip manufacturer

bet on an NVIDIA ARM chip

by

Microsoft has been looking for a convincing answer for a long time to a question that the PC has been dragging on for years: whether Windows on ARM can be something more than autonomy, silence and light productivity. He Surface Laptop Ultra It is born precisely there. The company describes it as its most powerful Surface Laptop yet and ties it to more ambitious workloads, from creation to local AI, leaning on NVIDIA RTX Spark as a great novelty. Microsoft’s move comes after a move that did not go unnoticed. Apple announced in 2020 the transition from the Mac to Apple silicon and materialized it that same year with the M1the company’s first own chip for its computers. With it, performance per watt, unified memory and vertical integration became an advantage that was difficult to ignore. That pressure no longer comes only from the high range. He MacBook Neowith a starting price of 699 euros, came to compete with affordable Chromebooks and Windows laptops. In parallel, Microsoft has tried to organize its own response with the Copilot+ PCa more solid category than its previous attempts with ARMbut still far from becoming a bet that has paid off in popularity and, consequently, in market share. A laptop to demonstrate an idea: Windows on ARM wants to go beyond autonomy The file that Microsoft has advanced It has two layers. The first is in performance: NVIDIA Blackwell RTX GPU, up to 128 GB of unified memory, CUDA and 1 petaflop of AI compute (based on theoretical FP4 TOPS with sparsity). The second is in everything you see and touch: a touch screen mini-LED 15-inch PixelSense Ultra, up to 2,000 nits of brightness HDR maximum, 262 pixels per inch and the largest haptic touchpad that Microsoft has included in a Surface. There is also something almost surprising in 2026: HDMI, USB-C, USB-A, SD and headphone ports without depending on adapters. When Microsoft talks about unified memory it doesn’t mean putting a lot of RAM in a laptop. The important thing is that this memory can serve as a common room for CPU and GPUso that the workloads are distributed according to what each task needs, although it eliminates the possibility of later extensions. The other half of the announcement is in what has not yet appeared. Microsoft has not detailed the price, the exact launch date, the final configurations, the specific processor, the exact GPU or the autonomy expressed in hours. One detail: the company itself remembers in the fine print that the Surface Laptop Ultra is a preliminary product, that its characteristics may change and that the sale will depend on the certifications or regulatory approvals of each country or region. Microsoft doesn’t get to this point from scratch either. The Copilot+ PCs of 2024 had already served to organize a new stage of Windows on ARM, with processors capable of surpassing the 40 TOPS of NPUlocal AI functions and equipment from partners such as Acer, ASUS, Dell, HP, Lenovo and Samsung, in addition to the Surface themselves. However, the available data invites us to put the growth in context. The Registerciting Context, placed Copilot+ PCs at only 9% of AI PCs distributed in Europe in the second quarter of 2025. The figure showed progress, but also a still small base. That’s where NVIDIA changes the type of conversation. The first wave of Windows on ARM relied heavily on efficiency, NPU and autonomy, but the Surface Laptop Ultra wants to add an ingredient that was missing from that story: graphical muscle and compatibility with a well-established professional ecosystem. CUDA It is not a minor word for anyone who works with AI, rendering or professional software supported by the NVIDIA ecosystem, because it has been part of many workflows for years. Microsoft is also introducing optimized Prism, native creative app support for ARM, and agreements with partners in gaming and anti-cheat systems. What we have before us seems like a very interesting proposal. The announcement invites us to take it seriously, because the Redmond firm has made its position quite clear: it wants this Surface to be proof that this new stage of Windows on ARM can go much further. But it is still incomplete to pass sentence.to. We’ll have to get our hands on it to see how it performs, how it behaves and how it works in real life. Images | Microsoft In Xataka | Apple has shown the weaknesses of laptops with the MacBook Neo. And the manufacturers are getting their act together

The advanced chip business is growing so fast that it cannot keep up

by

ASML, the Netherlands-based company that makes the most advanced integrated circuit production machines, had planned to hire 600 new employees in Taiwan this year. Finally she was forced to revise upwards your hiring plan. In 2026 they will arrive at their facilities on this Asian island 1,000 new additions. Grace Wang, the vice president and general manager of ASML in Taiwan, has declared that this change has been brought about by the insatiable demand for chips for artificial intelligence (AI). ASML does not manufacture semiconductors, but its equipment extreme ultraviolet photolithography (UVE) are being used by TSMC, SK Hynix, Samsung, Intel and Micron to produce the advanced integrated circuits that data centers demand. Especially CPU, GPU and HBM type DRAM memories (High Bandwidth Memory or high bandwidth memory). In fact, this company alone occupies the first link in the global chip manufacturing chain because it is the only one that produces EUV lithography machines. Be that as it may, Grace Wang’s declaration of intent responds to an unappealable reality: Taiwan is the industrial heart of this Dutch company. ASML manufactures components on this island and assembles UVE lithography equipment which it subsequently delivers to its local customers. These operations are also carried out in the Netherlands, but there are two compelling reasons why Taiwan is enormously relevant to ASML’s business: its best customer and its biggest focus on its global customers reside there. TSMC is ASML’s largest customer A determining factor that is promoting ASML’s expansion in Taiwan is its close relationship with TSMC, the largest manufacturer of integrated circuits of the planet. The operations of this company on the island currently generate 8.3 billion eurosa quarter of ASML’s global revenue. And much of that money comes from the coffers of TSMC, which is building new advanced semiconductor production plants in Taiwan, Japan, Germany and the US. ASML is building a facility in New Taipei that costs about $954 million. However, ASML’s Taiwan branch is not just hiring more staff (it currently has 4,500 employees on this island); is also building a new facility in New Taipei that costs about 954 million dollars. Their plan is for this plant to begin operating before the end of 2026 and to house about 2,000 employees in its initial phase. We still don’t know for sure what this factory will do, but it will probably combine component production, EUV machine assembly, and technical support to customers, primarily TSMC. ASML’s infrastructure in Taiwan is distributed between two cities with very specialized functions. Linkou is responsible for reconditioning chip manufacturing equipment, producing grating manipulators for deep ultraviolet (DVP) machines, and cleaning UVE collectors. Tainan, however, serves as a large global customer service center. And in a few months, as we have just seen, the New Taipei plant will be ready. The future of ASML is promising even though US sanctions They prevent it from selling its most sophisticated machines to its Chinese customers. Image | ASML More information | DigiTimes Asia In Xataka | The chip of the future comes from Japan: it is 1,000 times faster than current semiconductors and does not heat up

the IMEC chip laboratory has manufactured the first qubit with ASML’s High-NA machine

by

Manufacture a qubit, the physical device that implements the minimum unit of information in the quantum computersit is not at all a piece of cake. There are several types: superconductors, ion traps, neutral atoms or ions implanted in macromolecules, among other variants. Not all of them are equally complexbut all are difficult to produce and manipulate. In fact, the ideal is to be able to manufacture them on a large scale in order to make possible the arrival of quantum machines equipped with many more qubits than the current ones. The first step in this direction was taken by Intel and QuTech, the research institute specialized in quantum computing that belongs to the Technical University of Delft, in the Netherlands. At the end of March 2024 they announced that they had managed to produce the first qubit industrially and using the same processes and technology that is currently used for manufacture semiconductors. However, it is now IMEC (Interuniversity Microelectronics Center), the most experienced laboratory in the development of new integration and nanotechnology technologies that we have in Europe, which has signed a very important milestone: has managed to manufacture a qubit using extreme ultraviolet (UVE) and high aperture (High-NA) photolithography equipment from ASML. Currently this is integrated circuit manufacturing machine most advanced that exists. Caressing the dream of industrial manufacturing of qubits for quantum machines IMEC’s ​​main laboratory resides in Leuven, Belgium and has collaborated closely with ASML for more than four decades. Thanks to this collaboration you have access to the most advanced lithography equipment of the Netherlands company. The qubit produced using ASML’s High-NA equipment is a silicon quantum dot spin type. These qubits are very interesting because they are considered the most promising candidates for industrial scaling. In fact, as IMEC assuresthey are known as “the qubits of industry.” IMEC has shown that the manufacturing of these qubits is largely compatible with the production of CMOS chips The really relevant news is that IMEC has demonstrated that the manufacturing of these qubits is largely compatible with the production of integrated circuits using CMOS technology (Complementary Metal-Oxide-Semiconductor or complementary metal oxide semiconductor). And therefore it is possible manufacture them in conventional semiconductor plants. An important note: CMOS is the transistor manufacturing technology behind virtually all modern chips. Sofie Beyne, the director of this project at IMEC, maintains that “We can leverage decades of semiconductor innovation and repurpose the entire silicon scale-up ecosystem, taking quantum devices beyond laboratory experiments into large-scale, fabricatable systems. This is where silicon-based qubits have a clear advantage.” Experts who research in the field of quantum computing they are convinced that having machines with millions of qubits will lead to the arrival of error correction technology, which is the holy grail of these computers. Broadly speaking, silicon quantum dot spin qubits confine an electron within a silicon nanostructure, so that the spin state of the trapped electron is used to store quantum information. This architecture requires that the spaces between the different doors be minimal in order to reduce environmental noise and minimize errors. Be that as it may, what is really important is that IMEC has managed to manufacture a network of qubits with spaces of just 6 nm. Thanks to the nanoscale of this component, millions of qubits could theoretically be integrated into a single chip. Image | IMEC More information | IMEC In Xataka | China has reached one of the holy grails of quantum physics. So says Peter Zoller, father of quantum computers

This is how India prepares its next big leap in the chip industry

by

When we think about making chips, India is not usually the first country that comes to mind. We think of Taiwan, South Korea, the United States or China, but India often appears in another box: that of software, technological services or the assembly of electronic products. Precisely for this reason this movement is interesting. We are witnessing firsthand how a huge country seeks to advance towards one of the deepest and most difficult parts of the technological chain. The announcement is supported by a very specific alliance: Tata Electronics and ASML have signed a memorandum of understanding for the future Dholera semiconductor plant in Gujarat. The facility, which the release presents as India’s first 300mm commercial factory, has a planned investment of 11 billion dollars and will be aimed at producing semiconductors for a wide range of sectors. The agreement, Reuters points outwas signed on May 16 during Narendra Modi’s visit to the Netherlands, also attended by Dutch Prime Minister Rob Jetten. India’s commitment to chips is beginning to cease to be a promise The news comes accompanied by some interesting technical data that is worth analyzing carefully. The 300 mm figure may sound like a measurement of the chip, but it actually talks about the silicon wafer on which the semiconductors are manufactured, indicating the diameter of that circular surface, not the size of the transistors or the size of the final chips. It is a standard industrial platform for large-scale production, because it allows working with many units on the same wafer before cutting them and taking them to the next phases of the process. Furthermore, the statement places the Tata collaboration with PSMCPowerchip Semiconductor Manufacturing Corporation, as the gateway to a technology portfolio that includes 28nm, 40nm, 55nm, 90nm and 110nm. That list helps land the project much better than the 300 mm label, because it tells us which process families the plant plans to handle. In practice, we talk about analog and logic chips for automotive, mobile devices, consumer electronics, connectivity, IoT, embedded memory and other industrial uses. This is precisely where the Dutch company occupies a central position in the semiconductor industry. Their contribution does not simply consist of “setting up machines”, but rather supporting the start-up of the factory with lithography tools and solutions, the phase that allows the circuit patterns to be transferred to the silicon wafers. ASML also speaks training local talent, supply chain resilience and R&D infrastructure. It is a way to accompany Tata not only in the purchase of equipment, but in the industrial learning necessary for the plant to scale. ASML is known for its most sophisticated machines, but that doesn’t mean every deal of yours involves EUV either High-NA EUV. In this case, the available information speaks of lithography solutions for a process-oriented plant that they do not belong to the most advanced frontier. With these data, and in the absence of ASML detailing what specific systems it will supply, the most prudent reading is to place the project in the field of lithography DUV and the support ecosystem that allows stable production. TWINSCAN NXT:1980Di, one of the DUV machines that would fit with a factory like Tata’s This agreement comes at an interesting time for the Dutch company. ASML continues to occupy a central position in the sector, but the adoption of its most advanced technology is not being uniform among its large clients: TSMC, for example, continues to evaluate High-NA EUV, although for now it prefers to rely on current EUVs and process improvements rather than assume the cost and complexity of the new generation. Its executive Kevin Zhang summed it up with a very clear phrase: “I like the technology, but not the price.” In this context, an alliance like Tata’s reminds us of something important: the Dutch firm not only stands out for its most extreme machines, but also for its ability to accompany factories in different stages of industrial maturity. Reuters places the agreement in a broader context than that of a single factory. India has pledged billions of dollars in subsidies to attract semiconductor plants and related manufacturing, with eight projects underway, while Modi has encouraged Dutch companies to invest in areas such as semiconductors, renewable energy, digital technologies and health. At the same time, Dutch companies in the sector look for new markets and more geographical diversification in a scenario marked by export controls and trade restrictions linked to technological rivalry between the United States and China. As we can see, the move does not put India, at least for now, at the top of the semiconductor industry, but it does bring it closer to manufacturing chips at scale with proven processes and sustained demand. That difference matters. The industry does not only live on the most advanced semiconductors, but also on a huge base of components present in devices that surround us, such as everyday electronics and the automotive industry, as well as those that play in another field, such as industrial systems. Images | Tata Electronics | ASML In Xataka | Manufacturing 60 machines a year may not seem like much. In practice, those of the European ASML are setting the pace of AI

Your main competitor in chip manufacturing is your greatest ally

by

Approximately 30% of Intel chips It is manufactured outside of its semiconductor plants. And most of these ICs are produced by TSMC. This Taiwanese company is the largest chip manufacturer on the planetand therefore it is also Intel’s main competitor in the market for custom semiconductor manufacturing for third parties. Despite this, Intel is deeply dependent on TSMC. Currently, the latter company manufactures Intel’s most advanced integrated circuits in those nodes where the company led by Lip-Bu Tan has not yet reached optimal wafer performance. The CEO of Intel has confirmed recently that his company is also outsourcing the production of those chips that have frequent demand peaks, and which, therefore, is not able to manufacture in sufficient quantities to satisfy the needs of its customers. TSMC is producing Intel’s entire Lunar Lake line, and most of Arrow Lake as well. In addition, Intel is one of the first clients of this Taiwanese company with access to 2nm node latest generation. In fact, TSMC has already started production testing of the compute tile of Nova Lake at its Hsinchu plant, and large-scale manufacturing should begin before the end of 2026. During Intel’s latest financial results presentation, Tan confirmed something indisputable: TSMC is an essential partner of your company. TSMC also depends on Intel Intel’s current dependence on TSMC is a consequence of the delay accumulated for years in the development of its manufacturing nodes, before the arrival of Lip-Bu Tan. In fact, the delays in the jump to 10 nmand later the 7nmcontributed to TSMC overtaking Intel from a technological point of view. In addition, dependence on Intel is also based on the need to cover short-term demand. The dependence on Intel is based on the need to cover short-term demand This scenario clearly reflects that Intel and TSMC are competitors. But they are also partners. In fact, few people know that Intel sells chip manufacturing equipment to TSMC. Pat Gelsinger confirmed itthe former CEO of Intel, at the end of 2024 during a meeting with his investors: “TSMC is an impressive company. They serve their customers well, and they serve us well. Lunar Lake would not exist without TSMC (…) But we also supply them with some of our advanced equipment. “It is a complex relationship that is important for Intel, for TSMC and for the entire industry,” Gelsinger explained. It is clear that the former CEO of Intel wanted to convey positivity with this statement, and, to the extent possible, defend his management. We all know how it ended. Be that as it may, there is no doubt about one thing: Intel’s dependence on TSMC is not temporary; It is structural. It is the result of years of technological delays, is maintained by TSMC’s superiority in mature advanced nodes and has been consolidated as part of a strategy that prioritizes flexibility over in-house production. Image | Xataka In Xataka | Bad news for Intel and Europe: construction of Germany chip factory will be delayed until 2029 or 2030 In Xataka | TSMC promised them very happy with their new factory in Arizona. I wasn’t aware of the nightmare I was facing.

More advanced chip factories are being built in China and Taiwan than anywhere else. It’s only good for them

by

According to SEMI, an international organization that looks after the interests of the electronics and integrated circuit industries, only six of the 64 new factories of semiconductors that are going to come into operation in Asia before 2029 will reside in Southeast Asia. The remaining 58 They will be located in China and Taiwan. These two countries have compelling reasons to strengthen its chip industry and develop its integrated circuit production capacity. It is essential for China to set up new plants equipped with cutting-edge photolithography equipment. And that is precisely what SMIC, Hua Hong Semiconductor and other Chinese chipmakers are doing. Currently this nation is limited by the difficulty of going beyond 7 nm without being able to use the extreme ultraviolet lithography (VVE) of ASML. Even so, Huali Microelectronics, the division of Hua Hong Semiconductor specialized in manufacturing chips for third parties, is preparing to start the production of 7nm integrated circuits at its Shanghai plant. Taiwan also needs to expand its semiconductor industry, although its motives are very different from China’s. The two largest Taiwanese integrated circuit manufacturers, TSMC and UMCthey need to develop more cutting-edge plants in order to satisfy the growing needs of their customers. TSMC’s 2 and 3 nm nodes in particular cannot cope, so it is essential for this company to expand its production capacity in the midst of the boom in data centers for data applications. artificial intelligence (AI). SEMI is concerned about the vulnerabilities of the chip industry Ajit Manocha, the executive director of SEMI, assures that “we want to see more centers emerge in related countries. We want more plants to be established to reduce the risk derived from vulnerabilities.” What worries the spokesperson of this organization is that the geopolitical tensions maintained by the US, China and Taiwan end up threatening the integrated circuit factories that reside in these last two countries. TSMC’s in Taiwan are especially sensitive to a possible conflict with China due to the undoubted strategic importance that they have not only for Taiwan, but also for the US and its allies. Malaysia, Singapore, Vietnam and Thailand are candidates to host new cutting-edge chip plants Malaysia, Singapore, Vietnam and Thailand are strong candidates to host new cutting-edge chip manufacturing plants. In fact, Several centers already reside in Malaysia Intel’s advanced packaging and verification software. However, Manocha You are also concerned about other types of vulnerabilities. The most critical of all is the shortage of critical minerals, as well as bromine and helium, two fundamental gases in chip manufacturing processes. What is happening with helium in particular is very worrying. This gas is a byproduct of natural gas processing, and its price skyrocketed in March shortly after the war that the US, Israel and Iran have been fighting since then began because Qatar was forced to stop production of liquefied natural gas. In the current unstable scenario, SEMI argues that Southeast Asian countries should aim to build more semiconductor manufacturing plants over the next decade to help the sector diversify and reduce supply risks. Image | TSMC More information | Reuters In Xataka | The US’s problem in the AI ​​and humanoid race is not China: it is all of Asia and it is greatly disadvantaged

The chip crisis is leaving no stone unturned. Motherboards seemed untouchable, but their time has come

by

The RAM memory crisis is no longer a RAM memory crisis. Emulating what happened at the end of 2020, we are immersed in a new component crisis that, unlike that of five years ago, has not been caused by a combination of factors but by something very specific, the AI ​​industry. It is very difficult to buy any component with a NAND chip at a fair price and it is something that It is affecting all devices. The PC was already affected, but now the four largest motherboard manufacturers predict the worst. A shipment contraction of almost 30% on motherboards. Focused away from consumption. To understand why the crisis is impacting motherboards when, a priori, they could continue to be produced at the normal price, you have to look a little further. Nvidia and AMD are fully focused on the artificial intelligence segment, pausing their consumer GPU renewal plans for 2060. For example, the RTX 50 Super series neither is nor expected for this year and already speaks of some RTX 60 that would be released by 2028. Intel, for its part, also recently confirmed that consumer processors were taking a back seat in its priorities, since They were going to focus on the Xeon for servers and data centers. It is a strategy aligned with the great objective: become the great American foundry and sneak into the conversation they dominate TSMC and, at a good distance, Samsung. Update paused. With three manufacturers of the three key components having the focus away from the user and with the three main memory manufacturers (Samsung, Micron and SK Hynix) focused on memory for AI, what had to happen is happening: not only is buying new parts for a PC very expensive, but sometimes there is no stock, and the enthusiastic user who renews a PC every generation has no reason to do so, even if he had the deepest pockets in the world. That is why they are putting these PC updates on hold, clinging to current devices that they will have to get more out of because the market, directly, is broken. motherboards. And if PCs cannot be built and the companies that build computers themselves have already reported that they are having problems due to the price of RAM and storage, the foundation of the computer stops making sense. That’s where motherboards come in as those ‘foundations’. As we read in Tom’s Hardwarethe four main ones in the market (Taiwanese, too) are shipping units well below expectations. According to the media, ASRock will ship 37% fewer boards, Asus 33% less, MSI 24% less and Gigabyte 22% less. In total, the big four in this segment will ship 28% fewer units than they moved last year, which will push prices up for components that had not yet suffered the blow. In fact, the fact that the four companies are reviewing its shipping and sales predictions for this period could cause a parallel crisis. In a hypothetical scenario (because at the rate we are going it is very difficult for the crisis to be resolved in two days), if tomorrow it is possible to buy RAM and SSD memory at their normal price and people start building PCs again, what would be missing would be motherboards, so the shortage would cause price spikes. No end in sight. But hey, I already say that it is a hypothetical scenario because everything indicates that the NAND chip crisis is not even close to being resolved in the short term. The hyperscalers have the vast majority of the team unemployed full time, but they continue to demand new platforms to continue developing AI. That is why the entire segment has turned its head to look only at a market that is guaranteeing them an unprecedented peak in income. And no, don’t think that Asus, MSI, ASRock or Gigabyte are having a hard time, since they point that manufacturers are also offsetting declining profits in the user segment with growth in data center platforms. In the end, although ‘rare’, they are computers that need motherboards. When will the storm pass? Namely. There are sources that point to 2027 to start seeing green shoots, but others go to 2030 and Nvidia, which in the end has some hand in this mess, considers that there are seven or eight years of wild investment left. Image | Sitraka, Luis Gonzalez In Xataka | There is a shortage of RAM because of AI. That will make your next console much more expensive.

DeepSeek V4 has given China the boost it needs against the US. Four chip makers are the big winners

by

DeepSeek V4 It is the catalyst China needed. This model of artificial intelligence (AI) developed by the quantitative hedge fund specialized in trading algorithmic High-Flyer has been designed natively to live with Chinese chips. This is exactly the strategy that the Chinese government supports in response to the pressure that the US is putting on China. The Administration led by Donald Trump prevents the most powerful GPUs from Nvidia, AMD or Cerebras from reaching this Asian country. And Beijing has decided to do without them. The challenge facing the Chinese government is that it is much easier to set this goal than to put it into practice. This is the scenario in which DeepSeek V4 has emerged as the asset that China needs. And its arrival has led, for the first time, to several Chinese AI chip designers achieving something that until now had only been within the reach of Nvidia: guaranteeing full compatibility with the latest High-Flyer AI model from day 0. A great opportunity for Huawei, Cambricon, Moore Threads and Hygon DeepSeek V4 has marked a turning point. Its adoption in China is likely to be very notable, which has caused AI chip designers to compete among themselves to ensure full compatibility with this model from the moment it arrives. None of them wants to miss the opportunity to grow in the largest market on the planet if we stick to the most relevant indicators, such as purchasing power parity or the volume of population with the capacity to consume. Huawei is one of the companies that benefited most from the arrival of DeepSeek V4 Huawei will surely be one of the companies that will benefit the most from the arrival of DeepSeek V4. And its entire portfolio of GPUs for AI is compatible with this model. Nevertheless, your Ascend 950PR chip has been established as the main inference solution. A note before moving forward: inference is broadly speaking the computational process carried out by language models with the purpose of generating responses that correspond to the requests they receive. China’s three largest internet groups (Alibaba, ByteDance and Tencent) have placed orders for several hundred thousand Ascend 950PR processors following the launch of DeepSeek V4, according to Reuters. However, Huawei is not the only Chinese company that has won the lottery with the arrival of this AI model. Cambricon Technologiesthe Chen brothers’ company, has already completed the adaptation to the framework open source vLLM inference framework and has published the code on GitHub. Besides, Moore Threads has worked closely with the Beijing Artificial Intelligence Academy to run DeepSeek V4 on its MTT S5000 card using the FlagOS software stack. And Hygon has carried out a deep optimization of this model in its DCU platform with the purpose of consolidating its hardware as an attractive option for industrial use. The competitiveness of DeepSeek V4 outside of China is unclear because is less capable than its more advanced American competitors, but its future within the borders of its home country appears to be guaranteed. Image | Huawei More information | SCMP In Xataka | The US’s problem in the AI ​​and humanoid race is not China: it is all of Asia and it is greatly disadvantaged

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.